You are on page 1of 454

Internal Combustion Engine Modeling

Dr. D. N. Malkhede
Ph.D. (IIT Bombay)
Professor in Mechanical Engineering
Preface

It gives me immense pleasure to present this book for teachers, researchers and
students. This book is organized as a road map for researching, arguing, and
composing a literature review. Doing a literature review is a complex task for even
the most advanced researcher. To become successful at this craft, researchers need
many skills. The purpose of this book is to gather various internal combustion
engine models and their use for research at the predictive stages.
I congratulate the students of FY MTech (Automotive Technology) for their
contribution in making this event successful. I hope that this book will be useful
not only to students but also to the serving engineers in the automotive field.

Dr. D. N. Malkhede

Ph.D. (IIT Bombay)


IC ENGINE MODELING

Teaching Scheme Examination Scheme

Lectures: 3 hrs/Week Internal 40 marks

Tutorial: 2 hrs/Week End Sem Exam 60 marks

Fundamentals: Governing equations, Equilibrium charts of combustion chemistry, Chemical


reaction rates, Approaches of modeling, Model building and integration methods. Gas exchange
through valves, engine and porting geometry, exhaust gas recirculation, valve lift curves.

Thermodynamic Combustion Models of Engines: Single zone models, premixed and diffusive
combustion models, combustion heat release using Wiebe function, wall heat transfer
correlations, ignition delay, internal energy estimations, two-zone model, applications of heat
release analysis.

Modeling of Charging System: Constant-pressure and pulse turbocharging, compressor and


turbine maps, charge air cooler.

Fuel Spray Behavior: Fuel injection, overall spray structure, fuel atomization, spray
penetration, droplet size distribution, spray evaporation models, thick spray models, droplet
turbulence-interactions, droplet impingement on walls.

Mathematical Models of SI Engines: Simulation of Otto cycle at full throttle, part throttle and
supercharged conditions, progressive combustion, Autoignition Modeling, single zone models,
multi-zone models and mass burning rate estimation, SI engine with stratified charge. Friction in
pumping, in piston assembly, bearings and valve train etc. Friction estimation for warm and the
warm-up engines.
Text Books

1. Internal Combustion Engine Fundamentals, John B Heywood, McGraw-Hill, 1988.


2. Internal Combustion Engine Modeling, J.I. Ramos, Hemisphere Publishing Corporation,
1989.
3. Modeling Engine Spray and Combustion Processes, G. Stiesch, Springer Verlag, 2003.

Reference Books:

4. Turbocharging the Internal Combustion Engine, N. Watson and M.S. Janota, John Wiley &
Sons, New York, 1982.
5. Simulating Combustion: Simulation of combustion and pollutant formation for engine,
Gnter P. Merker, Christian Schwarz, Gunnar Stiesch, Frank Otto, Springer, 2008.
6. Introduction to Modeling and Control of IC Engine Systems, Guzzella Lino, Springer Verlag,
2004.
7. Internal Combustion Engines, R.S. Benson and N.D. Whitehouse, Volumes 1 and 2,
Pergamon Press, Inc. 1979.
8. The Thermodynamics and Gas Dynamics of Internal Combustion Engines, R.S. Benson,
Volume I and II, Edited by J.H. Horlock and D.E. Winterbone, Clarendon Press, Oxford,
1982. .
9. Thermodynamic analysis of combustion engines, Ashley, S, Campbell, John Wiley and
Sons, 1980.
10. Combustion Modeling in Reciprocating Engines, J. N. Mattavi and C. A. Amann, Plenum
press 1980.
11. Theory of Engine Manifold Design, D. E. Winterbone and R.J. Pearson, SAE, 2000.
12. Design Techniques for Engine Manifolds, D. E. Winterbone and R.J. Pearson, SAE, 1999.
13. Design and Simulation of Four-Stroke Engines, G. P. Blair, SAE, 1999.
14. Automotive Control Systems for Engine, Driveline and Vehicle, Uwe Kiencke and Lars
Nielsen, 2e, Springer, 2005.
15. Bosch Handbook
16. Modelling Diesel Combustion, Lakshminarayanan, P. A., Aghav, Yoghesh V., Mechanical
Engineering Series, Springer, 2010.

END SEM EXAMINATION: 60 Marks

INTERNAL EVALUATION: 40 MARKS

PROJECT REPORT

Marks 40
Stage I 15

Stage II 20

Software 05

How to do a Project? Project is individuals project: Project will be allocated to every


individual and to be prepared by individual. Should be submitted by individual and should be
presented by individual and should get evaluated by individual.

Based on MATLAB:

An algorithm is to be studied from books suggested. Prepare the notes based on literature.
Prepare flow chart. Prepare a program. It should be a MATLAB function like a black box
comprising of an algorithm. Function may have some inputs and some outputs.

Pre-Requisite:

Student must have passed IC Engine Course

Should have an Engine

Everybody should be able to work on MATLAB with ease

Engine: Dismantle the engine; take measurements of components, All those measurements
needed for simulating the engine.

Assignment using Software: AVL Boost and Ricardo Wave softwares are with us and you will
be expected to submit the assigned task.
Index

1. Introduction
2. Engine Thermodynamics
3. S. I. Engine Combustion
4. C. I. Engine Combustion
5. Turbocharger
6. Heat Transfer
7. Friction
8. Flow
9. Emission
Engine combustion modeling An Introduction

Modeling, in science and engineering, may be generally regarded as the process

of describing the physical phenomena in a particular system with the help of

mathematical equations (subject to reasonable assumptions) and solving the same to

understand more about the nature of such phenomena. Usually, engineering models help

in designing better devices by understanding more about the fundamental physical

processes occurring therein. Engine modeling activities, at least in recent decades, have

largely been concentrated in the direction of designing better performing engines with

lower emissions. In this regard, modeling of engine combustion processes assumes

importance.

The various engine combustion models that have been developed to date may be

grouped into three categories:

1. Zero dimensional models

2. Quasi-dimensional models

3. Multi-dimensional models

In the above classification, although the level of detail and proximity to physical

reality increases as one proceeds downward, so does the complexity of creating and using

those models.

Zero dimensional models are the simplest and most suitable to observe the effects

of empirical variations in the engine operating parameters on overall heat release rates1/

1
The phrase heat release rate is a misnomer, since heat is essentially a form of energy in transit and
cannot be released from a substance, say a fuel. The correct phrase could be either rate of conversion of
chemical energy to thermal energy or simply energy release rate. However, since heat release is
widely used in engine literature and is semantically compact, it will be retained here.

1
cylinder pressure schedules. These models are zero dimensional in the sense that they do

not involve any consideration of the flow field dimensions.

Zero dimensional models are further sub-divided into:

1. Single zone models

2. Two zone models

3. Multi-zone models

In single zone models, the working fluid in the engine is assumed to be a

thermodynamic system, which undergoes energy and/or mass exchange with the

surroundings and the energy released during the combustion process is obtained by

applying the first law of thermodynamics to the system.

In two zone models, the working fluid is imagined to consist of two zones, an

unburned zone and a burned zone. These zones are actually two distinct thermodynamic

systems with energy and mass interactions between themselves and their common

surroundings, the cylinder walls. The mass-burning rate (or the cylinder pressure), as a

function of crank angle, is then numerically computed by solving the simplified equations

resulting from applying the first law to the two zones.

A brief note is in order when referring to the nature of single zone and two zone

models. These models have been traditionally used in two different directions (shown in

Figure 1):

1. In one way, both these models have been used to predict the in-cylinder pressure

as a function of crank angle from an assumed energy release or mass burned

profile (as a function of crank angle).

2
2. Another use of these models lies in determining the energy release/mass burning

rate as a function of crank angle from experimentally obtained in-cylinder

pressure data.

Multi-zone models take this form of analysis one step further by considering

energy and mass balances over several zones, thus obtaining results that are closer to

reality.

Assumptions for a typical two zone model

1. The burned and unburned zones are ideal gases of different properties.

2. The unburned zone is assumed to consist of a premixed fuel-air mixture. Though this

may not be exact for diesel combustion, it is more realistic for SI engine combustion.

3. The characteristic gas constants of the burned and unburned zones do not vary much

with temperature and pressure; or if any such variations exist, they can be suitably

modeled using explicit relationships between gas constants and properties (T, P, etc.)

4. No heat transfer occurs from the burned to the unburned zone and vice versa.

5. Enthalpy associated with injected fuel is usually not significant and hence ignored.

6. Crevice losses may be significant but are not included.

7. Spatially averaged instantaneous heat transfer rates are adequate to estimate heat

transfer to the cylinder walls.

8. Instantaneous pressure in both the zones is the same since the flame is a deflagration

combustion wave.

9. The work required to transfer fluid from the unburned zone to the burned zone is

negligible.

3
ASSUMED HEAT RELEASE SCHEDULE CYLINDER PRESSURE

Assumed heat In-cylinder Performance


release (or mass pressure is parameters for
burned) profiles, determined from the particular
such as Wiebes the heat release engine under
function. rate consideration

EXPERIMENTAL PRESSURE DATA HEAT RELEASE SCHEDULE

Heat release/mass Heat release and


Experimental in- burning rate performance
cylinder pressure obtained from parameters
data pressure data

Figure 1. The directions followed by different variations of the single and two
zone models for different purposes.

4
mfuel

CONTROL
VOLUME
(single zone) Qht

mfuel

Burned zones

Unburned zones Qht

Figure 2. Schematics of the single zone (a) and two zone (b) combustion
models depicting the associated energy and mass transfers

5
Single zone Model (to find heat release rates from measured pressure)

Applying the First Law of Thermodynamics to the system (single zone):

Qch Qht W dm f dU
+ hf = (1)
d d d d d

where:

Qch
: Apparent rate of chemical energy (or heat) release
d

Qht
: Rate of heat transfer out of the system
d

W
: Rate of work transfer out of the system
d

dm f
h f : Rate of enthalpy inflow with the fuel
d

dU
: Rate of change of internal energy of the system
d

All the energy rates are expressed with respect to the crank angle .

The sum of the work transfer term and the internal energy change of the working

fluid term may be expressed in terms of P and V alone as follows:

W dU dV d
+ =P + (mcvT ) (2)
d d d d

where

m : Total mass of in-cylinder gases (the system) in kg

cv : Specific heat of the working fluid at constant volume in kJ/kg. K

6
From the ideal gas equation,

PV = mRT (3)

where

R : Characteristic gas constant of the cylinder gases in kJ/kg. K

Upon logarithmic differentiation, the above equation becomes

dP dV dm dR dT
+ = + +
P V m R T

However, considering the fact that the mass in the control volume remains constant

(neglecting crevice losses and quantity of fuel injected) and also assuming that the gas

constant remains constant throughout the combustion process, we get

dP dV dT
+ = (4)
P V T

dT
Substituting for m (from Equation 3) and (from Equation 4) in Equation 2, we get
T

dP dV
+
dW dU dV PV P V
+ =P + c v
d d d R d

Rearranging and simplifying,

dW dU cv dV cv dP
+ = 1 + P + V (5)
d d R d R d

Now for an ideal gas,

c p cv = R ( 6)
cp
= (7)
cv

7
where

c p : Specific heat of the working fluid at constant pressure in kJ/kg. K

: Ratio of specific heats at constant pressure and constant volume

Using Equations (6) and (7) and simplifying, Equation (5) becomes

dW dU dV 1 dP
+ = P + V (8)
d d 1 d 1 d

Substituting Equation (8) in Equation (1), neglecting the fuel enthalpy term, and

rearranging,

dQn dQch dQht dV 1 dP


= = P + V (9)
d d d 1 d 1 d

where

dQn
: Net apparent rate of heat release
d

Two zone Model (to find heat release rates from measured pressure)

In order to perform parametric simulation studies on engine combustion, a simple

thermodynamic model, which incorporates two zones, viz. an unburned zone and a

burned zone is desirable. Naturally, it is important to understand the suitability (or lack

thereof) of the two zone model to predict heat release rates in diesel and spark-ignited

engine combustion. Whereas in spark-ignited engines relatively well-defined

(continuous) flame propagation might occur, diesel engine combustion might consist of

several distinct burned and unburned regions scattered throughout the cylinder, as shown

in Figure 2. Despite the differences in the nature of combustion, the two zone model

8
would still be applicable for both diesel and spark-ignited engine combustion since it is

basically a zero dimensional model that considers only unburned and burned zones

without any consideration of the spatial location of such zones. The basic idea behind the

two-zone model is to utilize the conservation of mass and energy (First Law of

Thermodynamics) and also the ideal gas equations in obtaining an apparent rate of heat

release curve or apparent mass burned fraction curve. The adjective apparent should

be stressed in order to realize the fact that whatever be the heat release or mass burned

fraction obtained indirectly from measuring in-cylinder pressure, its accuracy is limited

by both the assumptions of the model as well as the accuracy of the pressure data.

The conservation of mass inside the cylinder is

m = mu + mb (10)

where

m : Total mass of charge inside the cylinder (= mfuel + mair)

mu : Mass of unburned charge

mv : Mass of burned charge

Also, since the mass inside the cylinder is assumed to be constant in any given engine

cycle,

. .
mb = mu (11)

Since the unburned and burned zones together constitute the total cylinder charge volume

at any instant, we have

V = Vu + Vb (12)

where

9
V : Total volume of charge inside the cylinder

Vu : Volume of unburned charge

Vb : Volume of burned charge

The ideal gas equations for the unburned and burned zones are

PVu = mu RuTu (13)


PVb = mb RuTb (14)

The generic form of the first law of thermodynamics applied to a control volume is

Q W . . dEcv
+ mh mh = (15)
dt dt in out dt

Now, considering the above expression for the unburned zone in terms of the respective

rates of the energy terms with crankangle,

. . . .
Qht ,u PVu + mu hu = mu uu (16)

Similarly, applying the first law (Equation 14) for the burned zone and remembering that

the total energy input into the burned zone is due to the chemical energy (released by the

fuel when it burns), we get

. . . . .
Qch Qht ,b P Vb mu hu = mbub (17)

The governing equations for the two zone model are Equations (10-14) and (16-17).

These equations are solved simultaneously each time (or crank angle) step to determine

the unknown quantities Tu, Tb, Vu, Vb,, mu, mb, and heat release rate from the known

cylinder pressure P and total volume V.

10
Copyrighted Material
Lecture-2
Kinematic Analysis of Internal Combustion Engine with Offset
Prof. D. N. Malkhede
Professor in Mechanical Engineering,
Governm ent College of Engineering, Pune
Em ail: dnm .m ech@coep.ac.in

Theengine transforms the reciprocating movement of the piston into the rotating
movement of the crankshaft. The piston reverses its movement at the top dead center (TDC) and
at the bottom dead center (BDC). At both of these dead point positions, the speed of the piston is
zero, while the acceleration is at the maximum. Between the top dead center and the underside of
the cylinder head, the compression volume Vc.

Fig.1. Assembly of reciprocating engine

Fig.1. shows kinematics ofa crankshaft drive with crossing, in which the longitudinal crankshaft
axis does not intersect with the piston wrist pin axis, but rather is displaced by the length e.
For the piston path s( ) , it follows from Fig.1:

s( ) c3 c2 r cos( )

Where,
e
sin( )
r 1
e
sin 1
r 1

c1 e r sin( )

c2 l 2 c12
c3 ( r l ) 2 e 2

Final equation can be written,

s (r l )2 e2 l 2 e r sin( ) r cos( )
2

The piston speed relation is given by,


ds re r sin( ) cos( )
r sin( )
d l 2 e r sin( )
2

With the definition of the cylinder volume,



V ( ) Vcomp D 2 s( )
4
Follows for the alteration of cylinder volume,
dV ds
D2
d 4 d

With the eccentric rod relation s r / l , it follows finally for the limiting case e = 0

1
s r [1 cos( )] [1 2s sin2 ( )]
s
And

ds s sin(2 )
rsin( )
d 2 [1 2s sin2 ( )]

For small s , the expression under the root in corresponding to

2s 4
[1 2s sin 2 ( )] 1 sin 2 ( ) s sin 4 ( ) ......
2 8
s = 0.25 already becomes smaller than 0.00048 and can thus be neglected as a rule. With the
help of trigonometric transformations, one finally obtains for thepiston path
s
1 cos( ) s (1 cos(2 ))
r 4
With the angular velocity ,
d

dt
One obtains for the piston speed
ds ds d ds

dt d dt d

The expression

ds
r sin( ) s (sin(2 ))
dt 2

And for the piston acceleration

d 2 s d 2 s d
22
2 d s

dt 2 d 2 dt d 2

Finally,

d 2s
2
r 2 cos( ) s (cos(2 ))
dt
4/4/2015

OVERVIEW OF THE PRESENTATION


CONTROL-ORIENTED MODELING AND Introduction
CONTROL OF TURBOCHARGED DIESEL Various Modeling Approaches
Control-Oriented Modeling Approach
ENGINE Nonlinear Model Development
Simulation Results
BY
DR. Dileep N. Malkhede Linearization of the Model
Comparison: Linearised with Nonlinear Model
Design of control system using
Classical Controllers
State feedback Controllers
Conclusions

Department of Mechanical Engineering


College of Engineering, Pune

INTRODUCTION INTRODUCTION
Diesel engines: A major source of power
Automotive Conventionally speed is governed by Centrifugal Governors
Marine
Inflexible
Generator sets
Its high thermal efficiency Compared to gasoline engines Only proportional type
Lower emissions

Drawbacks Need: Microprocessor controls


Low power density Stringent norms
Lower engine operating speeds
Fuel economy
Rough running and smoky exhaust
It has no natural top speed Emissions
Engine runs away without a proper control: Self-destruction Faster response
High speed fluctuations: fluctuating loads Flexible
Increased soot / wear and tear
Vibrations: affecting fuel economy

Issue therefore is to detect fluctuations in the speed and


Regulate fuel quantity by varying rack and
Maintain the engine speed within strict limits

Steps in Controller Design MODELING APPROACHES


Choice of domain: (Independent variable)
Crank angle-based (Instantaneous values at each oCA)
Engine controller design typically include Time-based
Although difficult to judge the suitability
Model Development: Control-Oriented Former approach is suggested for EGR/VGT control
Time domain approach is suggested for speed control
Controller Design Process
Controller Hardware Design Quasi Steady Method
Whole system is divided into different subsystems
Each subsystem is modeled by its steady state
characteristics
Large engine data is required
Manifold volumes are neglected: no mass accumulation is
considered
Algebraic relations: fast but engine specific

1
4/4/2015

LITERATURE: MODELING OBSERVER DESIGN


Filling and Emptying Method
Based on Mass Balance Where the states are not measurable/not
Volumes are not neglected. accessible/expensive
Mass accumulation (Pressure rise) is considered.
System model is embedded to act as an observer
States are computed by the observer
Pressure rise is modeled by ODEs
In IC Engines both manifold pressures fluctuate very
Mean Value Models rapidly
Average values over 3-5 revolutions are taken: Very difficult to measure: Expensive
Neglecting cyclic variations.
Continuous Time Based Models.
Spatial variations are neglected.
As if there are infinite cylinders

PROBLEM FORMULATION
Fourth order nonlinear model development
4 States: Engine speed is the output
Fuel Rack (Controlling input)
Propeller Pitch (disturbance)
11 Op. Points are Identified
Model is Linearised: Transfer Function Models / State-Space
Controller Design
Classical Controllers : Reduced order
MATHEMATICAL MODEL
PI Controller-pole placement
With and without anti-wind up
PID Controller based on ITAE optimal criterion
State Feedback Controllers: Full Order Model
State Feedback Regulator Design
Single input
Dual input
Speed Tracking Application
Partial State-Feedback Controller
Reduced Order Observer-Based Controller
Sensitivity Analysis

ENGINE SCHEMATIC ENGINE MODEL


Engine is divided in four major components
Prime-Mover shaft dynamics
Engine torque
Load torque
2 Turbocharger shaft dynamics
Compressor torque
Turbine torque
Inlet Manifold dynamics
Mass flow of air from Compressor to the intake manifold
Mass of air inducted in the engine
Exhaust Manifold dynamics
4 Mass of gas flow from engine to exhaust manifold
1 Mass of gas flow to turbine

Shaft Gas
Dynamics Flows
Notations Used
M Torque (N.m) Inlet Manifold Exhaust Manifold
3 Prime-Mover Turbo-Charger
T Temperature (K) Mass & Internal Energy Reservoirs
Torque Balance
Eulers Eq. of rigid body rotation Emptying and Filling Approach

2
4/4/2015

DEVELOPING SUBMODELS ESTIMATION OF ENGINE TORQUE


Indicated Engine Torque, Mi

Indicated Efficiency
PARAMETERS
State Variables Controlling Input Disturbance
e , tc , pin, pex h
Torques Engine Me= f (h, pin , e ) Ml = (e , )
at shafts Turbocharger Mt = f(tc, e, pin, pex, h) Mc = (pin, tc)
Mass Inlet Manifold mc = (pin , tc ) mie = f (pin, e )
Flows Exhaust Manifold mee = f (pin , e, h) mt = f (pin, h, tc, e, pex)

ENGINE PUMPING RATE EQUATION VOLUMETRIC EFFICIENCY

FUEL PUMPING RATE

ENGINE INDICATED TORQUE FRICTION TORQUE ESTIMATION

3
4/4/2015

LOAD TORQUE COMPRESSOR TORQUE

COMPRESSOR MAP Courtesy: Cummins India Ltd. Pune


TURBINE TORQUE CALCULATIONS

Turbocharger
speeds

ESTIMATION OF EXHAUST GAS TEMPERATURE a

ANALYTICAL EXPRESSIONS OF THE


MODEL

4
4/4/2015

PRIME-MOVER MODEL TURBOCHARGER MODEL

COMPRESSOR TORQUE ESTIMATION INLET MANIFOLD MODEL

EXHAUST MANIFOLD MODEL

SIMULATION OF OPEN LOOP STEP


RESPONSES

5
4/4/2015

NONLINEAR SIMULINK MODEL-OVERALL SIMULATION RESULTS - I

Overall Model

Simulation results

SIMULATION RESULTS - II SIMULATION RESULTS - III

SIMULATION RESULTS: FULL & 50% LOAD

MODEL LINEARIZATION

6
4/4/2015

IDENTIFICATION OF OPERATING POINTS LINEARIZATION: TURBOCHARGER MODEL

TURBOCHARGER EQUATION STATE SPACE MODEL

Partial differentials are obtained using symbolic toolbox of MATLAB.


Coefficients are evaluated at all 11 operating points

COMPARISON: LINEARISED & NONLINEAR

7
Copyrighted Material

Calculations of Internal Energy using Justi's Method

Prof. D. N. Malkhede
Professor in Mechanical Engineering,
Governm ent College of Engineering, Pune
Em ail: dnm .m ech@coep.ac.in

ABSTRACT
A thermodynamic system such as IC Engines is represented by its thermodynamic
properties. Total internal energy of the system is one of the most important properties needed for
thermodynamic modeling IC Engine system. The internal energy of IC Engine depends on parameters
temperature, pressure and air/fuel ratio. As these properties changes during 0-720o of CA in a cycle, the
internal energy of engine is continuously changing according to change in temperature, pressure &
air/fuel ratio of the working medium in the engine. Therefore it is essential to study the change in internal
energy with respect to temperature, pressure & air/fuel ratio. There are two popular approaches which
uses simple and empirical and still a more accurate approaches. One is Justi's correlation and the other
one is Zacharia's correlation.

The Justi's Method:


Investigations are carried out in order to describe the internal energy and enthalpy of combustion gasses
in terms of the quantities of temperature, pressure, and gas composition. The gas composition is thereby
expressed as the so-called air-fuel ratio. In describing diesel engine combustion gasses, this procedure is
correct as long as one stays within the super-stoichiometric range ( > 1). The composition of the
combustion air is assumed to be constant; a varying water content cannot be taken into consideration.
Fuel mixtures and alternative fuels cannot be represented with these methods. The relation is valid,
strictly speaking, only for a fixed C-H-ratio. This is however very similar to diesel and gasoline fuel. The
dissociation of the gas can also not be considered. Under dissociation we understand the change in
otherwise constant equilibrium constants of chemical reactions at very high temperatures, e.g. over 2,000
K. Since a single-zone combustion chamber model is used as a rule only for calculating the cylinder
pressure and the caloric mean temperature and with that for further thermodynamic quantities like
performance, efficiency, and heat flows, it is usually totally sufficient to represent the internal energy as a
function of temperature and the air-fuel ratio, for which Justi (1938) gives the following function
A Simple and widely used empirical function, established by Justi assumes that
the cylinder contents behave as an ideal gas. Therefore, the specific internal energy depends on
temperature and composition only. For further simplification the composition is not expressed in
terms of the mass fraction of all species in the mixture but in terms of reciprocal equivalence
ratio ? =1/. This implies that the ratio of C- to H- atoms of the fuel must not be changed, that
the global equivalence ratio has to be either stoichiometric or lean ( 1) to assure complete
combustion to CO2 and H2O and that the composition of combustion air has to be constant.
Thus, a varying water content of the intake air cannot be accounted for. However, since the C/H
ratio of gasoline and diesel fuel are similar adequate results have been obtained for both engines
types with following polynomial expression for the internal energy
? = ? (? , ? )

? ? .? ? .? ? ?
? (? , ? )=0.1445[1356.8 + ?489.6 + ? ? .? ? ? ?? ? ? ? ? ?10? ? + ?7.768 + ? .? ? ?? ? ? ? ? ? 10? ?
?
0.0485 ? ??
(0.0975 + ? .? ? )?? ? ? ? ? ? 10 ] ?? [ ]
??
? ??

Where, ? ? ? ? = 273.15 ?

INPUT VARIABLES Input parameters OUTPUT VARIABLE

Temperature, T (K), ? ? ? ? = 273.15 ? Specific internal


And energy, u (kJ/kg)
MATLAB PROGRAM:

clc
clear all
close all

% % u is the internal energy


% % T is the temperature
% % lambda is the relative air/fuel ratio
% % du/dt change in internal energy with respect to temperature.

% Justi's Correlation below


lambda = 1;
figure(1)
T = 0:100:3000;
u = 0.1445*(1356.8+ (489.6+(46.4/(lambda.^0.93))).*(T-273)*10.^(-
2)+(7.768+(3.36/(lambda.^0.8))).*((T-273).^2).*(10^(-4))-
(0.0975+(0.0485/(lambda.^0.75))).*((T-273).^3).*(10.^(-6)));
plot (T, u, '-b');
hold on

lambda = 2;
u = 0.1445*(1356.8+ (489.6+(46.4/(lambda.^0.93))).*(T-273)*10.^(-
2)+(7.768+(3.36/(lambda.^0.8))).*((T-273).^2).*(10^(-4))-
(0.0975+(0.0485/(lambda.^0.75))).*((T-273).^3).*(10.^(-6)));
plot (T, u, '--r');

lambda = 3;
u = 0.1445*(1356.8+ (489.6+(46.4/(lambda.^0.93))).*(T-273)*10.^(-
2)+(7.768+(3.36/(lambda.^0.8))).*((T-273).^2).*(10^(-4))-
(0.0975+(0.0485/(lambda.^0.75))).*((T-273).^3).*(10.^(-6)));
plot (T,u,'-m')

lambda = 15;
u = 0.1445*(1356.8+ (489.6+(46.4/(lambda.^0.93))).*(T-273)*10.^(-
2)+(7.768+(3.36/(lambda.^0.8))).*((T-273).^2).*(10^(-4))-
(0.0975+(0.0485/(lambda.^0.75))).*((T-273).^3).*(10.^(-6)));
plot (T,u,'--k')

xlabel('Temp in K'), ylabel('Sp. Int Energy (kJ/kg)'), grid on


title('Internal Energy as a fn of Temp and Lambda')
legend('\lambda = 1', '\lambda = 2', '\lambda = 3', '\lambda = 15', 2)
% Calculations of dudT and dudlambda
syms T lambda % Requirement of a Symbolic ToolBox
% Following is the expression to be differentiated
u = 0.1445*(1356.8+ (489.6+(46.4/(lambda.^0.93))).*(T-273)*10.^(-
2)+(7.768+(3.36/(lambda.^0.8))).*((T-273).^2).*(10^(-4))-
(0.0975+(0.0485/(lambda^0.75))).*((T-273).^3).*(10.^(-6)));
dudT = diff(u, T);
% % Using above, following is what you get
dudT = (8381./(125000.*lambda.^(93./100)) + (289.*(2.*T -
546).*(84./(25.*lambda.^(4./5)) + 971/125))./20000000 -
(867.*(97./(2000.*lambda.^(3./4)) + 39./400).*(T - 273).^2)./2000000000 +
44217./62500);
% % Above dydt is plotted for T between 0-3000
T = 0:100:3000;

lambda = 1;
dudT1=(8381./(125000.*lambda.^(93./100)) + (289.*(2.*T -
546).*(84./(25.*lambda.^(4./5)) + 971/125))./20000000 -
(867.*(97./(2000.*lambda.^(3./4)) + 39./400).*(T - 273).^2)./2000000000 +
44217./62500);
lambda = 2;
dudT2=(8381./(125000.*lambda.^(93./100)) + (289.*(2.*T -
546).*(84./(25.*lambda.^(4./5)) + 971/125))./20000000 -
(867.*(97./(2000.*lambda.^(3./4)) + 39./400).*(T - 273).^2)./2000000000 +
44217./62500);
lambda =4;
dudT4=(8381/(125000.*lambda.^(93/100)) + (289.*(2.*T - 546).*(84/(25.*lambda.^(4/5)) +
971/125))/20000000 - (867.*(97/(2000.*lambda.^(3/4)) + 39/400).*(T -
273).^2)/2000000000 + 44217/62500);
lambda = 6;
dudT6=(8381/(125000.*lambda.^(93/100)) + (289.*(2.*T - 546).*(84/(25.*lambda.^(4/5)) +
971/125))/20000000 - (867.*(97/(2000.*lambda.^(3/4)) + 39/400).*(T -
273).^2)/2000000000 + 44217/62500);
lambda = 8;
dudT8=(8381/(125000.*lambda.^(93/100)) + (289.*(2.*T - 546).*(84/(25.*lambda.^(4/5)) +
971/125))/20000000 - (867.*(97/(2000.*lambda.^(3/4)) + 39/400).*(T -
273).^2)/2000000000 + 44217/62500);
figure(2)
plot(T,dudT1);
hold on
plot(T,dudT2, '--r');
plot(T, dudT4, '-.k')
plot(T, dudT6, '-m')
plot(T,dudT8, ':b')
xlabel('Temp in K'), grid on
ylabel('\partial u/ \partial T ')
title('\partial u/ \partial T as a fn of Temp and at few Lambda')
legend('\lambda = 1','\lambda = 2', '\lambda = 4', '\lambda = 6', '\lambda = 8', 2)

figure(3)
lambda = 1 : 0.2 : 15;
T = 500;
dudT500=(8381./(125000.*lambda.^(93./100)) + (289.*(2.*T -
546).*(84./(25.*lambda.^(4./5)) + 971/125))./20000000 -
(867.*(97./(2000.*lambda.^(3./4)) + 39./400).*(T - 273).^2)./2000000000 +
44217./62500);
T = 1000;
dudT1000=(8381./(125000.*lambda.^(93./100)) + (289.*(2.*T -
546).*(84./(25.*lambda.^(4./5)) + 971/125))./20000000 -
(867.*(97./(2000.*lambda.^(3./4)) + 39./400).*(T - 273).^2)./2000000000 +
44217./62500);
T = 1500;
dudT1500=(8381./(125000.*lambda.^(93./100)) + (289.*(2.*T -
546).*(84./(25.*lambda.^(4./5)) + 971/125))./20000000 -
(867.*(97./(2000.*lambda.^(3./4)) + 39./400).*(T - 273).^2)./2000000000 +
44217./62500);

plot(lambda, dudT500)
hold on
plot(lambda, dudT1000, 'r')
plot(lambda,dudT1500, 'm')
xlabel('lambda'), ylabel('\partial u/ \partial T '), grid on
title('\partial u/ \partial T as a fn of Lambda at few Temperatures')
legend('Temp = 500', 'Temp = 1000', 'Temp = 1500', 1)
axis([0 15 0.5 1.15])

figure(4)
syms T lambda
dudlambda = diff(u,lambda);
% What you get is the following
dudlambda1=(84099.*(T - 273).^3)./(16000000000000.*lambda.^(7/4)) - (6069.*(T -
273).^2)./(156250000.*lambda.^(9/5)) - (779433.*(T -
273))./(12500000.*lambda.^(193/100));

T = 0:50:3000;
lambda = 1.0;
dudlambda_1 = (84099.*(T - 273).^3)./(16000000000000.*lambda.^(7/4)) - (6069.*(T -
273).^2)./(156250000.*lambda.^(9/5)) - (779433.*(T -
273))./(12500000.*lambda.^(193/100));
lambda = 2;
dudlambda_2 = (84099.*(T - 273).^3)./(16000000000000.*lambda.^(7/4)) - (6069.*(T -
273).^2)./(156250000.*lambda.^(9/5)) - (779433.*(T -
273))./(12500000.*lambda.^(193/100));
lambda = 3;
dudlambda_3 = (84099.*(T - 273).^3)./(16000000000000.*lambda.^(7/4)) - (6069.*(T -
273).^2)./(156250000.*lambda.^(9/5)) - (779433.*(T -
273))./(12500000.*lambda.^(193/100));
plot(T, dudlambda_1)
hold on
plot(T, dudlambda_2)
plot(T, dudlambda_3)
xlabel('Temperature K'), ylabel('\partial u / \partial \lambda'), grid on
legend('\lambda = 1.0', '\lambda = 2.0', '\lambda = 3.0', 3)
title('Variation of \partial u by \partial \lambda with Temp for few \lambda')
% axis([300 3000 0.5 1.1])

figure(5)
lambda = 1 : 0.2 : 15;
T = 500;
dudT500 = (84099.*(T - 273).^3)./(16000000000000.*lambda.^(7/4)) - (6069.*(T -
273).^2)./(156250000.*lambda.^(9/5)) - (779433.*(T -
273))./(12500000.*lambda.^(193/100));
T = 1000;
dudT1000 = (84099.*(T - 273).^3)./(16000000000000.*lambda.^(7/4)) - (6069.*(T -
273).^2)./(156250000.*lambda.^(9/5)) - (779433.*(T -
273))./(12500000.*lambda.^(193/100));
T =1500;
dudT1500 = (84099.*(T - 273).^3)./(16000000000000.*lambda.^(7/4)) - (6069.*(T -
273).^2)./(156250000.*lambda.^(9/5)) - (779433.*(T -
273))./(12500000.*lambda.^(193/100));
plot(lambda,dudT500), hold on, grid on
plot(lambda,dudT1000, 'r')
plot(lambda,dudT1500, 'k');
xlabel('\lambda')
ylabel('\partial u / \partial \lambda')
legend('Temp = 500', 'Temp = 1000', 'Temp = 1500', 4)
title('Variation of \partial u by \partial \lambda with \lambda for few Temps')
axis([0 15 -80 10 ])
Copyrighted Material

Calculations of Internal Energy using Zacharia's Method

Prof. D. N. Malkhede
Professor in Mechanical Engineering,
Governm ent College of Engineering, Pune
Em ail: dnm .m ech@coep.ac.in

This is in continuation with the previous lecture in which we had discussed about a method
(model) suggested by Justi for evaluating a specific internal energy of working medium in IC
engine. The Justi's correlation is empirical in nature as also the other one being presented the
model in this lecture. Justi's correlation relate internal energy with temperature and mixture
strength . Zacharia's correlation presented here suggests additional dependence on pressure.

Zacharias Correlation:-
Zacharias considered a pressure dependency to avoid the ideal gas assumption. Mixture
composition is approximated by the reciprocal equivalence ratio .
6

( , , )= 0 [ 2
1+ + () 1] [ ]
0

Where,
0 28.89758 0.0602
= 0.980665
= 000
=

= . 1 1 4 . 11 1 5
= . . 131
( ) = 3. 1 .
(1) = .131 3 .3 3
( )= . 1 .1 1
(3) = . 3 .
( )= . 1 + . 1 11
( )= . 1 3 . 3
( )= . 3 . 1 1 1

Zacharias method gives a better accuracy than Justis ideal gas assumption, it also
causes an increase in computational effort since an iterative solution of pressure and temperature
becomes necessary. Because of high temperatures generally encountered in combustion systems
the numerically simpler approach by Justi is still often used in thermodynamic models. Only in
engines with a high level of supercharging or turbo charging and with peak pressures in excess of
150 bar it is suggested to use pressure dependent model as real gas effects do become important.

Output variables Input variables Input parameter


Specific internal energy, u Universal gas constant, R (8.314
Temperature T (K)
(kJ/kg) kJ/kgK)
Equivalence ratio, Stoichiometric AFR
Pressure, P (bar)

Unit of Pressure P is not clearly given. So not sure


While plotting unit of p is taken as kPa, though.
REFERENCES:
1. Simulating Combustion Gunter P Merker Christian Schwarz Gunnar Stiesch Frank Otto
.
2. John B. Heywood. IC ENGINE
1

Analysis of Cylinder Pressure Data

Cylinder Pressure changes with crank angle as a result of change in volume,combustion


process, heat transfer to the cylinder walls flows into and out of the crevices regions and leakage.
The first two of this effects are largest . The effect of the volume change on the pressure can
readily be accounted for thus combustion rate information can be obtained from accurate
pressure data provided models for the remaining phenomenon can be developed at an
approximate level of an approximations.

Cylinder pressure can be measured with piezoelectric transducers. This type of


transducers contain the quartz crystals.One end of this crystal is exposed through a diaphragm to
the cylinder Pressure.As the cylinder pressure increases this crystal is compressed and generates
an electrical charge which is proportional to the pressure. A charge amplifier is then used to
produce an output voltage which is proportional to the charge or pressure.Accurate cylinder
pressure vs crank angle data can be obtained with the system provided following steps are taken .

1) The correct reference pressure is used to convert the measured pressure signal to
absolute pressure.
2) The pressure versus crank angle phasing is accurate within the 0.2?
3) The clearance volume is estimated with sufficient accuracy.
4) Transducer temperature variations due to the variation in wall heat flux during the
engine cycle are held to a minimum.

Log p vs Log v plots can be used to check the quality of cylinder pressure data . The first
three of the above requirements can be validated using the log p vs log v plots for a motored
engines . If effect of the thermal cycling are significant , the expansion curve on the log p vs log
v curve shows the excessive curvature.
2

Figure(1)

(a) Pressure Volume diagram (b) log p-log(? /? ? ??) plot at 1500 rev/min, imep = 513 kPa , =0.8,
? ? = 8.72, Propane Fuel.

Figure 1(a) and (b) shows pressure volume data from firing spark ignition engine on both
linear p-V and log p log V diagram . On the log p-log V diagram the compression process is a
straight line with the slope of 1.3 . The starting of the combustion can be identified by the
departure of the curve from the straight line. The end of combustion can be located
approximately in the same fashion; the expansion stroke following combustion is essentially
linear with the slope of 1.33 . Since both the compression of the unburned mixture prior to
combustion and expansion of the burned gases following the end of combustion are close to the
adiabatic isoentropic process , the observed behavior is as expected .The compression and
expansion behavior are more fitted with the polytropicbehaviour

? ? ? = ? ? ? ??? ? ? ---- (1)

The exponent n for the compression and expansion processes is 1.3 for conventional
fuels. It is comparable to the average value of ?? for the unburned mixture for the compression
process , but is larger than the ? ? for the burned gas mixture during heat loss to the combustion
chamber walls.
3

Log p-log V plots shown in the 1(b) approximately define the start and end of
combustion, but do not provide the burned mass fraction profile. Following equation could be
used to account for the effect of cylinder volume change on the pressure during combustion.
Assuming that the unburned gas filling the volume ?? at any crank angle during combustion has
been compressed polytropically by the advancing flame front , then the volume ? (? ,? ) it occupied
at time of spark is

?
? ? , ? = ? ? ( )? /? ---- (2)
? ?

Similarly, the burned gas behind the flame filling the volume ? ? would, at the end of
combustion, fill a volume ? ? ,? given by

?
? ? , ? = ? ? (? )? /? ---- (3)
?

The mass fraction burned ? ? is equal to 1- (? ? ,? /? ? ) and to ? ? ,? / ?? , where ? ? and ?? are


the total cylinder volumes at the time of spark and at the end of combustion , respectively. Since
= ?? + ? ? . Eqn (2) and (3) then give :

? ? /? ? ? 0? /? ? ?
?? =
? ? ? /? ?? ? ? ? /? ? ?

This method is widely used though it contains several approximations. Heat transfer
effecta are included only to extent that the polytropic exponent n in above equation differs from
? . The pressure rise due to combustion is proportional to the amount of fuel energy released
rather than the mass of mixture burned. Also the polytropic exponent not constant during
combustion. Selecting an appropriate value for n is major difficulty in applying this pressure data
analysis.

The effect of heat transfer , crevices and leackage can be explicitly incorporated in the analysis
by Heat Release approach. The major advantage of such type approach is that the pressure
changes can be related directly to the amount of fuel chemical energy released by combustion,
4

while retain the simplicity of treating the combustion chamber as the single zone. The first law
for the open system of the combustion chamber is ,

? ? ?? = ? ? ? + ? ? ? ? + ? ? + ? + ? ? ? ---- (5)

The change in sensible energy of the charge is separated from that , due to change in the
composition : the term on left side represent the energy released by the combustion. The work is
piston work and equal to ? . ? ? . ? ? ?? is heat transfer to the chamber walls . The mass flux term
represent the flow across the system boundary. In the absence of fuel injection it represents flow
into and across the crevices region.

The accuracy with which this energy balance can be made depends on how adequately
each term in above equation can be quantified. Assume that ? ? is given by mu(T) , where T is
the mean charge temperature and m is the mass within the system boundary, then ,

? ? ? = ? ?? (? ) + ? (? )? ? ---- (6)

Note that this mean temperature determined by the ideal gas law is close to the mass
average cylinder temperature during combustion since the molecular weights of the burned and
unburned gases are essentially same. Crevices effects can be usually modeled by flow into and
out of the single volume, with the gas in the crevice at substantially low temperature. Leackage
to the crankcase can be usually neglected. Then the above equation , on substitution for ? ? ? and
? ? ? becomes,

? ? ?? = ? ?? ? ? + (? ? )? ? ?? + ? ? ? + ? ? ? ? ---- (7)

Where,

?? ?? > 0 when flow is out of the cylinder into the crevices

?? ?? < 0 when is from the crevices to the cylinder

h is equivalent at cylinder conditions when ? ? ?? > 0 and at crevices conditions when ? ? ?? < 0.

Use of ideal gas law (neglect the change in the ideal gas constant R) then above equation will
give ,
5

? ?
? ? ?? = ? ?? ? ? ? ? + ? ? + 1? ? ? ? + (? ? + ?? ? )? ? ?? + ? ? ? ? ---- (8)
?

This equation can be used in various ways when heat and energy release term is
combined, with the heat transfer and crevices term, the combination termed net heat release. It is
equal to the first two terms on the right side of above equation, which together represent the
sensible energy change and the heat lost to the piston. While heat lost during combustion is the
small fraction of the fuel energy, the distribution of heat release and heat transfer with the crank
angle are different. Heat transfer becomes more important as the combustion process ends and
average gas temperatures peak. The net heat profile can be obtained by integrating the first two
terms in the above equation, normalized to give unity at its maximum value is often interpreted
as the burned mass fraction versus crank angle profile. Use of the above equation requires the
value of the ? ? /? . The ratio of specific heats ? decreases with increasing temperature and varies
with composition. As the mean charge temperature increase during compression and combustion
and then decrease during expansion ,? should vary. An approximate approach, modeling
? (T)with a linear function of temperature fitted to the appropriate curves with ? constant during
combustion, has been shown to adequate results.

MATLAB Program:

The following Matlab Programs can be used for the evaluation of p-V graph and Log p-
Log V graphs .

Program 1 : for the Pressure vs Volume data

p = [155.148 151.969 148.786 145.643 142.442 139.079 135.535 131.833


128.018 124.139 120.254 116.447 112.795 109.355 106.213 103.432 100.989
98.7708 96.6923 94.7172 92.8286 91.0179 89.2869 87.6291 86.0329 84.47
82.9521 81.4885 80.0848 78.7419 77.4584 76.2337 75.066 73.9475 72.857
71.797 70.7777 69.798 68.8574 67.9551 67.0893 66.2565 65.4331 64.6215
63.8296 63.0553 62.295 61.5463 60.8077 60.0789 59.3602 58.6523 57.9559
.
.
.
.There is two page values data.]
% Volume values as per given for the standard Engine
6

plot (v,p)% Graph of the pressure vs Volume will be plotted


xlabel ('Fraction of Maximum Volume')% label for the xaxis
ylabel ('Pressure(kPa)')% label for the yaxis

Program 2 : for the Log P vs Log V plot :

p = [155.148 151.969 148.786 145.643 142.442 139.079 135.535 131.833


128.018 124.139 120.254 116.447 112.795 109.355 106.213 103.432 100.989
98.7708 96.6923 94.7172 92.8286 91.0179 89.2869 87.6291 86.0329 84.47
82.9521 81.4885 80.0848 78.7419 77.4584 76.2337 75.066 73.9475 72.857
71.797 70.7777 69.798 68.8574 67.9551 67.0893 66.2565 65.4331 64.6215
63.8296 63.0553 62.295 61.5463 60.8077 60.0789 59.3602 58.6523 57.9559
57.2557 56.5667 55.9016 55.266 54.6631 54.0937 53.5579 53.056 52.5898
52.1629 51.7795 51.4441 51.1609 50.9338 50.7655 50.6464 50.5837 50.5764
50.617 50.6964 50.8065 50.9406 51.0933 51.2607 51.4405 51.6311 51.8312
52.039 52.2527 52.4715 52.695 52.9219 53.1507 53.3799 53.6088 53.8366
54.0624 54.2858 54.5066 54.7251 54.9423 55.1593 55.3777 55.5978 55.8169
56.0323 56.2431 56.4502 56.655 56.8591 57.0637 57.2695 57.4767 57.685
57.8944 58.1051 58.3169 58.5273 58.7408 58.9592 59.1819 59.4087 59.6394
59.8745 60.1147 60.3602 60.6114 60.8682 61.1312 61.3995 61.673 61.9515
.
.
.
.
.
There is two page values data.

]% Volume values as per given for the standard Engine


a = log10 (p) ;% calculate the log values of the pressure
b = log10 (v) ;% calculate the log values of the volume fractions
plot (b,a)% plot the graph for the logpvslogV
xlabel ('Fraction of Maxixmum Volume')
ylabel ('Pressure,kPa'
7

Outputs :
Thermodynamic Data Calculations

Gordon and McBride fitted curves to the tabulated JANAF data (and a
similar approach has been adopted for the Chemkin curves to give
expressions for the thermodynamic properties of the form.

The Matlab function airdata.m returns the curve coefficients a1 to a7 for


either the Gordon and McBride curves or / and the Chemkin curves over
two different temperature ranges:

1) 300 < T < 1000 K; and


2) 1000 < T < 5000 K.

Have a look at the Table at the Janaf Data below


How program works:
There are 10 rows in the returned matrix and each row provides the
seven coefficients (in ascending order) for a different species. The 10
species for which data is available are (in the following order):

CO2, CO, OH, and


H2O, H2, NO.
N2, H,
O2, O,

The options for airdata.m are listed in Table below, and the numerical
values for the curve coefficients can be found in the listing of
airdata.m.
The specific heat curves for the 10 species using the Gordon and
McBride and the Chemkin coefficients can be generated and Fig. xyz.

These figures indicate that differences between the two sets of


coefficients are relatively insignificant. For example, with the O2 curve,
the maximum difference between the values of cp for the two different
schemes is only 0.3% at around 3500 K.

By itself, this is a relatively small difference in engineering terms. When


it is recognized that

Table 2: Values of hof


and so using the Gordon and McBride and the Chemkin coefficients
compared with JANAF values.
Calculation of mole fraction

Chemical Equilibrium

It is a good approximation for performance estimates in engines to regard the burned gases
produced by the combustion of fuel and air as in chemical equilibrium. By this we mean that the
chemical reactions, by which individual species in the burned gases react together, produce and
remove each species at equal rates. No net change in species composition results.

For example, if the temperature of a mass of carbon dioxide gas in a vessel is increased
sufficiently, some of the CO 2 molecules dissociate into CO and O 2 molecules. If the mixture of

CO 2 , COand O 2 is in equilibrium, then CO 2 molecules


are dissociating into CO and Oz at the same rate as CO and Oz molecules are recombining in the
proportions required to satisfy the equation
1
CO + 02 = CO 2
2
In combustion products of hydrocarbon fuels, the major species present at low temperatures are
N 2 , H 2 O, CO 2 , and O 2 or CO and H 2 . At higher temperature (greater than about 2200 K), these
major species dissociate and react to form additional species in significant amounts. For
example, the adiabatic combustion of a stoichiometric mixture of a typical hydrocarbon fuel with
air produces products with species mole fractions of: N 2 - 0.7, H 2 O, CO 2 - 0.1, CO, OH, 0 2 ,

NO, H 2 -0.01 H 2 0 -0.001 and other species in lesser amounts.

The second law of thermodynamics defines the criterion for chemical equilibrium as follows.

Consider a system of chemically reacting substances undergoing a constant-pressure, constant-


temperature process. In the absence of shearwork (and electrical work, gravity, motion,
capillarity), the first law gives
Q=dH

The second law gives

1
Q =T dS

Combining these gives


dH- T dS=0

Since we are considering constant-temperature processes, this equation holds for


finite changes:
H-T S= G=0
Thus, reactions can only occur (at constant pressure and temperature) if G (=H - TS) for the
products is less than G for the reactants. Hence at equilibrium
G=0

Consider a reactive mixture of ideal gases. The reactant species M a ,M b , etc and product
speciesetc., are related by the general reaction whose stoichiometry is given by
V a M a +V b M b + ... = V i M i + V m M m +
This can be written as

i
Vi Mi = 0

where the Vi are the stoichiometric coefficients and by convention are positive for the product
species and negative for the reactant species.

Let an amount naofMa react with nbofMb, etc., and produce n, of M


nof M etc. These amounts are in proportion:
ni = vi n

The change in Gibbs free energy of a mixture of ideal gases, at constant .pressure and
temperature, as the composition changes is given by
(G)=
i
ui ni

2
where n i is the chnagein number of moles of species i and u is the chemical potential. The

chemical potential, an intensive property, is defined as


G
ui =
ni
It is equal in magnitude to the specific Gibbs free energy at a given temperature and pressure.
For an ideal gas
pi
u i = u i o *T + RT*ln
po

whereu i equals g i the standard specific Gibbs free energy of formation The standard state
pressure Po is usually one atmosphere.
Substitution in Eq. (3.36) gives, at equilibrium,

pi
(u i o *T + RT*ln ) n i =0
i po

We can divide by n i and rearrange, to obtain kp

Kpis the equilibrium constant at constant pressure:

It is obtained from the Gibbs free energy of the reaction which can be calculated from the Gibbs
free energy of formation of each species in the reaction,Itis a function of temperatUfe only.
In the JANAF tables,8 to simplify the calculation of equilibrium constants values of log 10 (K p )

the equilibrium constants offormation of one mole of each species from their elements in their
standard states, are tabulated against temperature.The equilibrium constant for a specific reaction
is then obtained via the relation.

log 10 (K p )=
i
v i log 10 (K p ) i

The coefficients are obtained by least-squares matching with thermodynamic property data from
the JANAF tables.
Usually two sets of coefficients are included for two adjacent temperature intervals (in the
NASA program these are 300 to 1000 K and 1000 to 5000 K)

3
In some equilibrium programs, the species to be included in the mixture must be specified as an
input to the calculation. In the NASA program, all allowable species are included in the
calculation, though species may be specifically omitted from consideration.

For each reactant composition and pair of thermodynamic state variables,


the program calculates and prints out the following:
1. Thermodynamic mixture properties
(obtained from the equilibrium composition and the appropriate gas mixture rule). p, T,
P, h, s, M, and a (sound speed)
2. Equilibrium composition:-Mole fractions of each species (which are present
insignificant amounts), Xi
Figure shows how the equilibrium composition of the products of combustion of isooctane-air
mixtures at selected temperatures and 30 atm pressure varies with the equivalence ratio. At low
temperatures, the products are N2 ,CO2 ,H20, and O2 for lean mixtures and N2' CO2 , H20, CO,
and H2 for rich mixtures. As temperature increases, the burned-gas mixture composition
becomes much more complex with dissociation products such as OH, 0, and H becoming
significant.

CHEMICAL EQUILIBRIUM USING EQUILIBRIUM CONSTANTS

This section presents a solution for the proportions of equilibrium combustion products based on
an equilibrium constant method applied by Olikara and Bonllan( 1975) to the gas phase products
of combustion of hydrocarbon fuels. Use of equilibrium constants is based on the minimization
of the Gibbs free energy of the gas. The equilibrium constant method is simpler than the
Lagrange multiplier approach when considering restricted species lists. However, the equilibrium
constant method requires that the equilibrium reactions, such the water-gas reaction.

Therefore, a more complete reaction calculation needs to be performed first to determine the
significant product species to include in the equilibrium constant analysis. Inspection shows that
if <3, the only species of importance

4
Because of dissociation arc 0, H, OH, and NO. The species list in Equation can be terminated at
i = 10; that is, we need to consider only 10 species. Let us consider the following reaction:

as
C H O + (3.76N 2 + O 2 ) n 1 CO 2 + n 2 H 2 O + n 3 N 2 + n 4 O 2 + n 5 CO +

n 6 H 2 + n 7 H + n 8 O + n 9 OH + n 10 NO

Atom balancing yields the following four equations:

C: = (n 1 +n 5 )N

H: =(2n 2 + 2n 6 +n 7 + n 9 )N

as
O: +2 = (2n 1 + n 2 + 2n 4 + n 8 + n 9 + n 10 )N

as
N: 2(3.76) = (2n 3 + n 10 )N

Where N is the total number of moles

We now introduce six gas phase equilibrium reactions. These reactions include the
Dissociation of hydrogen, oxygen, water, carbon dioxide , and equilibrium OH and NO

Sr no. A B C D E
1 0.432168E+00 -0.11246E+5 0.267269E+1 -0.745744E-4 0.242484E-8
2 0.310805E+00 -0.12954E+5 0.321779E+1 -0.738336E-4 0.344645E-8
3 -0.141784E+0 -0.213308E+4 0.853461 0.355015E-4 -0.310227E-8
4 0.150879E-01 -0.470959E+4 0.66096 0.272805E-5 -0.154444E-7
5 -0.752364E+0 0.12421E+1 -0.26028E+1 0.259556E-3 -0.162687E-7

5
6 -0.415302E-2 0.148627E+5 -0.475746E+1 0.124699E-3 -0.900227E-8

TABLE ; equilibrium constant curve-fit coefficient

formulation:
1 y7 P1 / 2
H 2 HK 1 = 1 / 2
2 y6

1 y P1 / 2
O 2 OK 2 = 8 1 / 2
2 y4

1 1 y9
H 2 + O 2 OH K3 = 1/ 2
2 2 y 4 y 61 / 2

1 1 y
O 2 + N 2 NO K4 = 10
1/ 2 1/ 2
2 2 y 4 3y

1 y2
H 2 + O 2 H 2 O K5 = 1/ 2
2 y 4 y6 P1 / 2

1 y
CO + O 2 CO 2 K6 = 1
1/ 2
2 y5 y 4 P1 / 2

The unit of pressure in the Sixe quations is the atmosphere. Olikotaand Borman(1975) have
curve fitted the equilibrium constants K,(7) to JANAFTable datafor 600 <T <4000 K. Their
expression are of the form

Bi
Log 10 K 6 (T) = A, ln(T/ 1000) + + C, + D,T+ ET 2
T

whereT is in Kelvin. The equilibrium constant curve-fit coefficients are listed in Table.

6
Given pressure p, temperature T, and equivalence ratio Equations willyield 11 equations for
the11 unknowns: the I0 unknown mole fractions Yi
And the unknown total product moles N. The set of 11equations are nonlinear and solved by
Newton Raphson iteration. With the product mole fraction composition known, one can proceed
to compute the thermodynamic properties of interest ,enthalpy ,entropy, specific volume. and
Internal energy.

7
Matlabprogram :-

1) Program for calculation of equilibrium constant

function mole(t)
% to calculate values of equilibrium constants at various temperature range
% input = t (in kelvin)
% t=1700;
% for ionization reaction of H and O
k1=10^((0.121343)*log(t/1000)+((0.017076e5)/t)+(-0.05451e1)+((-0.007408e-
4)*t)+((-0.102191e-8)*t^2))
% formation of OH (k2)
k2=10^((-0.141784)*log(t/1000)+((-0.213308e4)/t)+(0.853461)+((0.355015e-
4)*t)+((-0.310227e-8)*t^2))
% formation of NO (k3)
k3=10^((0.150879e-1)*log(t/1000)+((-0.470959e4)/t)+(0.646096)+((0.272805e-
5)*t)+((-0.154444e-8)*t^2))
% for water gas reation (k4)
k4=10^((-74.821098e-2)*log(t/1000)+((-0.024417e5)/t)+(-
0.21546e1)+((0.134857e-3)*t)+((0.73754e-8)*t^2))
formatshortG
end

2) Calculation of mole fraction

function mole(t)
% to calculate values of equilibrium constants at various temperature range
% input = t (in kelvin)
% t=1700;
% for ionization reaction of H and O
k1=10^((0.121343)*log(t/1000)+((0.017076e5)/t)+(-0.05451e1)+((-0.007408e-
4)*t)+((-0.102191e-8)*t^2))
% formation of OH (k2)
k2=10^((-0.141784)*log(t/1000)+((-0.213308e4)/t)+(0.853461)+((0.355015e-
4)*t)+((-0.310227e-8)*t^2))
% formation of NO (k3)

8
k3=10^((0.150879e-1)*log(t/1000)+((-0.470959e4)/t)+(0.646096)+((0.272805e-
5)*t)+((-0.154444e-8)*t^2))
% for water gas reation (k4)
k4=10^((-74.821098e-2)*log(t/1000)+((-0.024417e5)/t)+(-
0.21546e1)+((0.134857e-3)*t)+((0.73754e-8)*t^2))
formatshortG
end
% function mol(alpha,beta,t)
% inputs
% alpha=corbon content in fuel
% beta = hydrogen content in fuel
% t=temperature at which reaction is occuring
mole(t)
as=alpha+ beta/4
fi=0:0.01:1.4;
symsalphabetan1n2n3n4n5n6n7n8n9n10fi;
m=solve
('alpha=n1+n5','beta=2*n1+2*n6+n7+n9','2*(as./fi)=2*n1+n2+2*n4+n8+n9+n10','2*
3.76*(as./fi)=2*n3+n10','k1=(n7*(n4^0.5))/(n8*(n6^0.5)','k2=n9/(n4*n6)^0.5','
k3=n10/(n4*n3)^0.5','k4=n2*n5/n6*n1')
m= [m.n1 m.n2 m.n3 m.n4 m.n5 m.n6 m.n7 m.n8 m.n9 m.n10];
N=n1+n2+n3 +4+ n5+ n6+ n7+ n8+ n9+ n10;
mf1=n1/N
mf2=n2/N
mf3=n3/N
mf4=n4/N
mf5=n5/N
mf6=n6/N
mf7=n7/N
mf8=n8/N
mf9=n9/N
mf10=n10/N
% output
% mf1= mole fraction of co2
% mf2= mole fraction of h2o
% mf3= mole fraction of n2
% mf4= mole fraction of o2
% mf5= mole fraction of co

9
% mf6= mole fraction of h2
% mf7= mole fraction of h
% mf8= mole fraction of o
% mf9= mole fraction of oh
% mf10= mole fraction of no
plot(mf1,fi)
plot(mf2,fi)
plot(mf3,fi)
plot(mf4,fi)
plot(mf5,fi)
plot(mf6,fi)
plot(mf7,fi)
plot(mf8,fi)
plot(mf9,fi)
plot(mf10,fi)
end

References

1. I te al o ustio e gi e fu da e tals he wood , the o he ist of ai -fuel


i tu e , page o.- to .
2. IC e gi es fe guso , ge e al he i al e uili iu , page o - .

10
CHEMICAL EQUILIBRIUM USING EQUILIBRIUM CONSTANTS

This section presents a solution for the properties of equilibrium combustion products based
on an equilibrium constant method applied by Olikara and Borman (1975) to the gas phase
products of combustion of hydrocarbon fuels. Use of equilibrium constants is based on
minimization of the Gibbs free energy of the gas. The equilibrium constant method is simpler
than the Lagrange multiplier approach when considering restricted species lists. However, the
equilibrium constant method requires that the equilibrium reactions, the water-gas reaction
given by Equation 3.45, be specified.

Therefore, a more complete reaction calculation needs to be performed first to


determine the significant product species to include in the equilibrium constant analysis.
Inspection of Figures 3-2 and 3- shows that if < , the only species of importance because
of dissociation are O, H, OH, and NO. the species list in Equation 3.54 can be terminated at i
= 10; that is, we need to consider only 10 species. Let us consider the following reaction:

CH O Nb + (as/)*(O2 + 3.76 N2) n1CO2 + n2H2O + n3N2 + n4O2 n5CO + n6H2 +


n7H + n8O + n9OH + n10NO (3.71)

Atom balancing yields the following four equations:

C: = (y1 + y5) N
(3.72)

H: = (y2 + 2y6 +y7 + y9) N


(3.73)

O: + (as/) = (y1 + y2 + 2y4 + y5 + y8 + y9 + y10) N


(3.74)

N: + ((Z/.76as)/) = (y3 + y10) N


(3.75)

Where N is the total number of moles. By definition, the following can be written

We now introduce six gas phase equilibrium reactions. These reactions include the
dissociation of hydrogen, oxygen, water, carbon dioxide, and equilibrium OH and NO.
Table 3-6 Equilibrium Constant Curve-Fit Coefficients for Equation 3.83

Sr. A1 B1 C1 D1 E1
1 0.432168E + 00 -0.11246-1E + 0.267269E + 01 -0.745744E 0.242464E 08
2 0.310805E + 00 05 0.321779E + 01 04 0.344645E 08
3 -0.141784E + 00 -0.129540E + 05 0.853461E + 00 -0.738336E -0.310227E -08
4 0.150879E 01 -0.213308E + 04 0.646096E + 00 04 -0.154444E
5 -0.75236-1E + 00 -0.470959E + 04 -0.260286E + 01 0.355015E 04 08
6 -0.415302E - 02 0.124210E + 05 -0.475746E + 01 0.272805E 05 -0.162687E
0.148627E + 05 0.259556E 03 07
0.124699E - 03 -0.900227E - 08

Formation:

H2 = H K1 = (y7p1/2)/y61/2 (3.77)

O2 = O K2 = (y8P1/2)/y41/2 (3.78)

1/2 H2 + O2 = OH K3 = y9/(y41/2 y61/2) (3.79)

1/2O2 + N2 = NO K4 = y10/(y41/2y1/23) (3.80)

H2 + O2 = H2O K5 = y2/(y41/2y6P1/2) (3.81)

CO = O2 = CO2 K6 = y1/(y5y41/2P1/2) (3.82)

The unit of pressure in the six equations (Equations 3.77 to 3.82) is the atmosphere. Olikara
and Borman (1975) have curve fitted the equilibrium constants K1(t) to JANAF Table data
for 600<T<4000K. their expressions are of the form
Log10K1(T) = A1 in(T/1000) + B1/T + C1 + D1T + E1T2 (3.83)

Where T is in Kelvin. The equilibrium constant curve-fit coefficients are listed in table 3-6.
Given pressure P, temperature T, and equivalence ratio , Equations 3.72 to 3.82 will yield
11 equations for the 11 unknowns: the 10 unknown mole fractions y1 and the unknown total
product moles N. The set of 11 equations are nonlinear and solved by Newton-Raphson
iteration. With the product mole fraction composition known, one can proceed to compute the
thermodynamic properties of interest: enthalpy, entropy, specific volume, and internal
energy.

EXAMPLE 3.4 Equilibrium Combustion Mole Fraction

What are the mole mixture properties resulting from the combustion of a
gasoline mixture at a pressure of 3000kPa, temperature of 3000K, and a fuel-
air equivalence ratio of 1.2?

SOLUTION

The web pages contain an applet Equilibrium Combustion Solver that computes the product
mole fractions and properties for five fuels given the pressure, temperature, and the fuel-air
equivalence ratio. The fuels are gasoline (C7H17), diesel (C14.4H24.9), methanol (CH3OH),
methane (CH4), and nitromethane (CH3NO2).

Pressure = 3000

Temperature = 3000

Fuel Air equivalence ratio = 1:2

Press enter for combustion:

MOLE FRACTIONS: MIXTURE PROPERTIES

Mole Fraction CO2: 0.066 h (KJ/kg): 1336.3

Mole Fraction H2O: 0.138 u (KJ/kg): 411.5

Mole Fraction N2: 0.681 s (KJ/kgK): 9.416

Mole Fraction O2: 0.0033 v (m3/kg): 0.308


Mole Fraction CO: 0.0693

Mole Fraction H2: 0.0199 Molecular Weight: 26.97

Mole Fraction H: 0.0041 iterations: 86

Mole Fraction O: 0.0012 error: 0.0

Mole Fraction OH: 0.0095

Mole Fraction NO: 0.0058

The Equilibrium Combustion Solver applet can be used to compute general trends for fuel-air
combustion that are not immediately obvious. For example, the effect of temperature on
enthalpy of combustion products for three different equivalence ratios is shown in Figure 3-5
for the combustion of gasoline at a pressure of 101.3kPa. Note that the lowest value of
enthalpy occurs at a stoichiometric equivalence ratio, and as the equivalence ratio is made
lean or rich, below 2500K, the enthalpy increases. This behavior is also shown in Figure 3-6,
a plot of the enthalpy of the combustion products of methanol versus equivalence ratio at
pressures of 101kPa and 2000kPa. The enthalpy is minimum at near stoichiometric
conditions, as on either side of stoichiometric, the combustion is incomplete. If the mixture is
lean, there is an excess of unburnt oxygen. If the mixture is rich, there will be unburnt carbon
monoxide. A minimal value of enthalpy implies that the specific heaty of the combustion
products is also a minimum, which will maximize the adiabatic flame temperature.
1

Equivalence Ratio Determination from Exhaust Gases

Exhaust gas composition depends on the relative proportions of fuel and air fed to the engine,
fuel composition, and completeness of combustion. These relationships can be used to determine
the operating fuel/air equivalence ratio of an engine from a knowledge of its exhaust
composition. Composition of fuel can be represented as CnHmOr. The fuel is CnHmOr;the
oxidizer is air (O2 + 3.773N2). The products are CO2, H2O, CO, H2, O2, NOx, N2, unburned
hydrocarbons (unburned fuel and productsof partial fuel reaction), and soot particles (which are
mainly solid carbon).The amount of solid carbon present is usually sufficiently small (0.5
percent ofthe fuel mass) for it to be omitted from the analysis.

no 2
CnHmOr (O2 3.773N 2 ) n p ( X C a H b .Ca H b X CO .CO X O2 .O2 X N 2 N 2 X NO . XNO X NO2 .NO2 X H 2 O .H 2O X H 2 .H 2 )

(1)

Where is the measured equivalence ratio ((F/A)actual/(F/A)stoichiometric).


no2 is the number of O2 molecules required for complete combustion.

m r
n o2 n (2)
4 2
n p is the total number of moles of exhaust products, and Xiis the mole fraction of the ith

component.
There are seven unknowns which are , XH2, XH2O, XN2, np, a, b.

To solve these unknowns we use atomic balance for each equation.


Carbon Balance: n n p (aX CaHb X CO X CO2 ) (3)

Hydrogen Balance: m n p (b. X CaHb 2 X H 2O 2 X H 2 ) (4)

2nO2
Oxygen Balance: r n p ( X CO 2 X CO2 2 X NO 2 X O2 X H 2O ) (5)

7.546nO2
Nitrogen Balance: n p ( X NO 2 X N 2 ) (6)

Mole fractions add up to 1:
2

X CaHb X CO X H 2 X H 2O X N2 X NO X CO2 X O2 1 (7)

X CO . X H 2O
K (8)
X CO2 . X H 2

Where K is constant, values of 3.8 and 3.5 are commonly used for K.
The value of can be obtained by solving above equations using MATLAB.

For fuels comprised of carbon and hydrogen only, when all species are measured with the same
background moisture (wet, dry, or partially dry), the following expression based on the ratio of
measured and computed oxygen-containing species to measured carbon-containing species gives
the air/fuel ratio. It has been assumed that the unburned hydrocarbons have the same C/Hratio as
the fuel.

+ (NO 2 ) + (O 2 )]/[(HC) + (CO) + (CO 2 )]


A Mf CO H 2 O NO
( )actual = 4.773( )[(CO 2 ) + + +
F Mair 2 2 2
(9)

Where ( ) are molar concentrations (all with the same background moisture) in percent.
Mair = 28.96,
Mf = 12.01 + 1.008y (10)
Where y is the H/C ratio of the fuel.
(CO 2 ) + (CO)
(H 2 O) = 0.5y (11)
[CO/K(CO 2 ) ] + 1
Since nitrogen oxides collectively comprise less than 0.5 percent of the exhaust mixture, their
concentrations can be omitted with negligible error.
34.56(4 + y)
(A/F)stoichiometric = (12)
(12.011 + 1.008y)

(A/F)stoichiometric (F/A)actual
Equivalenc e ratio( ) = (13)
(A/F)actual (F/A)stoichiometric

Problem Definition:
3

To predict equivalence ratio from exhaust HC emissions, exhaust O2and from exhaust NO2.

Values for CO and CO2are predicted from following graph for calculation purpose.

Figure 1. Plot of molar fractions of exhaust products vs Equivalence ratio

MATLAB PROGRAM 1:
function [ phi ] = equivalence_ratio(CO, CO2, NO2, O2, HC) % Give exhaust
constituent's molar concentrations in % and y is H/C ratio
%y=2-2.25 for fuel compositions(gasolines and isooctane)
NO=0.05;
y=2.25;
K=3.8; %constant
u=CO+CO2;
v=CO/[K*CO2];
w=v+1;
x=u/w;
H2O=0.5*y*x;
Mf=12.01+(1.008*y);
Mair=28.96;
A=4.773*(Mair/Mf);
B=CO2+(CO/2)+(H2O)/2+(NO/2)+ NO2+O2;
C=HC + CO+ CO2;
4

D= B/C;
Air_Fuel_Actual= A*D;

E=34.56*(4+y);
Air_Fuel_Stio=E/Mf;
Eq=Air_Fuel_Stio/Air_Fuel_Actual;
fprintf('\nEquivalence ratio= %f', Eq);
end

Instructions to user to run program:


equivalence_ratio(CO, CO2, NO2, O2, HC) all inputs in % molar fractions.

Test inputs:
equivalence_ratio(2, 10, 1.2, 0, 16) Equivalence ratio= 2.347654
equivalence_ratio(2, 10, 1.2, 16, 0) Equivalence ratio= 0.541375

MATLAB PROGRAM 2
function [ Equivalence_ratio ] =equivalence_determination(NO2, O2, HC) % Give
exhaust constituent's mole concentrations(fractions) in %
%y=2-2.25 for fuel compositions(gasolines and isooctane)
NO=0.05;
% corresponding CO and CO2 molar fractions depending on O2
O2=[0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8
3.9 4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8
5.9 6 61 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 7];
CO=[9.78 9.33 9.22 4.81 2.07 0.921 0.771 0.586 0.511 0.511 0.4 0.398 0.35
0.324 0.3 0.285 0.285 0.284 0.284 0.283 0.255 0.245 0.244 0.243 0.243 0.242
0.242 0.23 0.215 0.204 0.22 0.239 0.238 0.238 0.238 0.238 0.237 0.236 0.25
0.28 0.308 0.27 0.232 0.232 0.232 0.232 0.232 0.231 0.202 0.193 0.21 0.229
0.229 0.229 0.227 0.227 0.226 0.2 0.188 0.171 0.15 0.12 0.09 0 0 0 0 0 0 0
0];
CO2=[8.49 8.74 8.82 11.3 12.9 13.4 13.5 13.5 13.4 13.4 13.3 13.2 13.2 13.2
13.1 13 13 13 13 12.9 12.8 12.7 12.7 12.6 12.6 12.6 12.6 12.6 12.5 12.5 12.4
12.3 12.3 12.2 12.1 12.1 12 12 12 11.9 11.8 11.7 11.6 11.6 11.6 11.5 11.5
5

11.4 11.4 11.3 11.3 11.2 11.12 11.05 11 10 10.9 10.8 10.8 10.7 10.6 10.5 10.4
10.4 10.3 10.3 10.1 10.1 10.1 10.1 10.1];

y=2.25;
K=3.8; %constant
u=CO+CO2; %formula==>
v=CO/[K*CO2];
w=v+1;
x=u/w;
H2O=0.5*y*x;
Mf=12.01+(1.008*y);
Mair=28.96;
A=4.773*(Mair/Mf);
B=CO2+(CO/2)+(H2O)/2+(NO/2)+ NO2+O2;
C=HC + CO+ CO2;
D= B/C;
Air_Fuel_Actual= A*D;

E=34.56*(4+y);

Air_Fuel_Stio=E/Mf;
Eq=Air_Fuel_Stio/Air_Fuel_Actual;
fprintf('\nEquivalence ratio= %f\n', Eq);
end

Instructions to user to run program:


File name is equivalence_determination(NO2, O2, HC).
In this molar concentrations are put in the form of % molar fractions
Test inputs:
equivalence_determination(NO2, O2, HC) O2 must fall between 0 to 7.0% for this analysis
equivalence_determination(1.2, 0.01, 12) Equivalence ratio= 1.570893

REFERENCES
Internal Combustion Engines Fundamentals, John B Heywood, Mc-Graw Hill Publication, pg
no. 145-151.
1

Mathematical Models for Property and Composition Calculations:

This section deals with relationships which model the composition and/or thermodynamic
properties of unburned and burned gas mixtures. The composition of gas mixtures have CO2,
H2O, CO, H2, O2, N2, OH, NO, O and H.
The models are based on polynomial curve fits to the thermodynamic data for each of these
species in the mixture and the assumptions that
(1) The unburned mixture is frozen in composition and
(2) The burned mixture is in equilibrium.
The approach used as the basis for representing JANAF table thermodynamic data in the
NASA equilibrium program is summarized here because it is consistent with the approach
used throughout to calculate unburned and burned mixture properties.
For each species i in its standard state at temperature T(K), the specific heat cp,i is
approximated by

c~p ,i
~ ai1 ai 2T ai 3T ai 4T ai 5T
2 3 4
(1)
R

The standard state enthalpy of species i is then given by


~
hi ai 2 a a a a
~ ai1 T i3 T 2 i 4 T 3 i5 T 4 i6 (2)
RT 2 3 4 5 T

The standard state entropy of species i at temperature T(K) and pressure 1 atm, is then
~
si ai 3 2 ai 4 3 ai 5 4
~ ai1 ln T ai 2T T T T ai 7 (3)
RT 3 4 5
2

Fig: Specific heat at constant pressure, cp/R, as a function of species CO2, H2O, O2, N2, H2,
CO. (From JANAF tables)

Values of the coefficients of aij for CO2, H2O, CO, H2, O2, N2, OH, NO, O, and H are given in
Table 1. Two temperature ranges are given. The 300 to 1000 K range is appropriate for
unburned mixture property calculations. The 1000 to 5000 K range is appropriate for burned
mixture property calculations. However, since the maximum temperature of 3000 K is highly
enough, plots for these properties within the range of 300 to 3000 K are prepared and
presented here and the figure gives values of cp/R for major species, CO2, H2O, H2, O2, N2,
and CO.

Unburned Mixtures

Polynomial functions for various fuels (in the vapor phase) have been fitted to the functional
form:
3

Af 5
c~p , f A f 1 A f 2 t A f 3t 2 A f 4 t 3 2 (4)
t

~ t2 t3 t 4 Af 5
h f Af 1 Af 2 Af 3 Af 4 Af 6 Af 8 (5)
2 3 4 t

Where t= T(K)/1000

Af6 is the constant for the datum of zero enthalpy for C, H2, O2 and N2 at 298.15 K. For a 0 K
datum, Af8 is added to Af6. For pure hydrocarbon compounds, the coefficients Afi were found
by fitting Eqs. (4) and (5) to data from Rossini. Values for relevant pure fuels are given in
Table 2. The units for cpf are cal/gmol K, and for hf are kcal/grnol.

Multicomponent fuel coefficients were determined as follows. Chemical analysis of the fuel
was performed to obtain the H/C ratio, average molecular weight, heating value, and the
weight percent of aromatics, olefins, and total paraffins (including cycloparaffins). The fuel
was then modeled as composed of a representative aromatic, olefin, and paraffin
hydrocarbon. From atomic conservation of hydrogen and carbon and the chemical analysis
results, component molar fractions and average carbon numbers can be determined. Table 2
gives values for the coefficients Af1 to Af8 for typical petroleum-based fuels. The units of the
coefficients give cpf and hf in cal/gmol.K and kcal/gmol, respectively, with t= T(K)/1000.
The thermodynamic properties of the unburned mixture can now be obtained. With the moles
of each species per mole O2, ni, determined from Table 3, and the mass of mixture per mole
O2, mRP determined from Table 4, the burned mixture properties are given by

1
c p ,u
mRP
n c~
i
i p ,i

1 ~
hu
mRP
n h i
i i

1
~ ~ ni ~

su ni si R ln nu R ln p
mRP
nu

4

where p is in atmospheres.
Figure (a) and (b), obtained with the above relations, show how cp,u and u (cp,u/cv,u) vary
with temperature, equivalence ratio, and burned gas fraction, for gasoline-air mixture.

Figure : (a) Figure: (b)


Specific heat at constant pressure of Ratio of specific heats, u= cp,u/cv,u
unburned gasoline, air, burned gas of unburned gasoline, air, burncd gas
mixtures as function of temperature, mixtures as function of temperature,
equivalence ratio, and burned gas equivalence ratio and burned gas fraction.
fraction. Units: kJ/kg mixture* K

Table no: 1
5

Table no: 2

Table no: 3
6

Table no: 4
7

Here,
= the molar N/O ratio (3.773 for air)
y = the molar H/C ratio of the fuel
= fuel/air equivalence ratio
ni = moles of species i per mole O2 reactant
xb= burned mass fraction

Steps for program:


The first program is for cp/R versus temperature for various species i.
1. For various species, values of the coefficients are obtained from the table no 1. For
temperature range of 300K to 1000K, the coefficients used are different and for
temperature range of 1000K to 3000K the coefficients are different.
2. The graph of cp/R v/s temperature is plotted accordingly for different species i.
3. And the corresponding values can be obtained using table 1 and table 2.
8

MATLAB Program:
First Part of the Program:
Assignment1_Part1

% This Program calculates cp/R vs T for various species


% For CO2
T= [300:1000];
x= (0.24008E+01)+(0.87351E-02*(T))+(-0.66071E-05*(T.^2))+...
(0.20022E-08*(T.^3))+(0.63274E-15*(T.^4)); %x=cp/R
plot(T,x)
hold on
T= [1000:3000];
x= (0.44608E+01)+(0.30982E-02*(T))+(-0.12393E-05*(T.^2))+(0.22741E-
09*(T.^3))+(-0.15526E-13*(T.^4));
plot(T,x, '-r')
text(T(end), x(end), 'CO_2')
% For H20
T= [300:1000];
x= (0.40701E+01)+(-0.11084E-02*(T))+(0.41521E-05*(T.^2))+(-0.296371E-
08*(T.^3))+(0.80702E-12*(T.^4));
plot(T,x)
T=[1000:3000];
x= (0.27168E+01)+(0.29451E-02*(T))+(-0.80224E-06*(T.^2))+(0.10227E-
09*(T.^3))+(-0.48472E-14*(T.^4));
plot(T,x, '-r')
text(T(end), x(end), 'H_2O')
% For CO
T= [300:1000];
9

x= (0.37101E+01)+(-0.16191E-02*(T))+(0.36924E-05*(T.^2))+(-0.20320E-
08*(T.^3))+(0.23953E-12*(T.^4));
plot(T,x)
T=[1000:3000];
x=(0.29841E+01)+(0.14891E-02*(T))+(-0.57900E-06*(T.^2))+(0.10365E-
09*(T.^3))+(-0.69354E-14*(T.^4));
plot(T,x,'-r')
text(T(end), x(end), 'CO')

% For H2
T= [300:1000];
x= (0.30574E+01)+(0.26765E-02*(T))+(-0.58099E-05*(T.^2))+(0.55210E-
08*(T.^3))+(-0.18123E-11*(T.^4));
plot(T,x)
T=[1000:3000];
x=(0.31002E+01)+(0.51119E-03*(T))+(0.52644E-07*(T.^2))+(-0.34910E-
10*(T.^3))+(0.36945E-14*(T.^4));
plot(T,x,'-r')
text(T(end), x(end), 'H_2')
% For O2
T=[300:1000];
x=(0.36256E+01)+(-0.18782E-02*(T))+(0.70555E-05*(T.^2))+(-0.67635E-
08*(T.^3))+(0.21556E-11*(T.^4));
plot(T,x)
T=[1000:3000];
x=(0.36220E+01)+(0.73618E-03*(T))+(-0.19652E-06*(T.^2))+(0.36202E-
10*(T.^3))+(-0.28946E-14*(T.^4));
plot(T,x,'-r')
text(T(end), x(end), 'O_2')
% For N2
T=[300:1000];
x=(0.36748E+01)+(-0.12082E-02*(T))+(0.23240E-05*(T.^2))+(-0.63218E-
09*(T.^3))+(-0.22577E-12*(T.^4));
plot(T,x)
T=[1000:3000];
x=(0.28963E+01)+(0.15155E-02*(T))+(-0.57235E-06*(T.^2))+(0.99807E-
10*(T.^3))+(-0.65224E-14*(T.^4));
plot(T,x, '-r')
text(T(end), x(end), 'N_2')
grid on
xlabel('Temperature K')
10

ylabel('c_p/R')

Output of the Program:

Fig no:1 Specific heat at constant pressure, cp/R as function of temperature for species
CO2, H2O, O2, N2, H2, and CO.
11

Second Part of the Program:


Assignment1_Part2

% This program calculates value of cp and plots it against temperature


% x1,x2,x3,x4,x5,x6 are the values of cp for the respective species
% ni= moles of each species per mole O2
% m_RP=mass of mixture per mole O2

clc
clear all
T=[300:1000];
%_________________________________________________________________________
%-----------------for phi=0------------%
% for O2
x1=1.987*(0.36256E+01)+(-0.18782E-02*(T))+(0.70555E-05*(T.^2))+(-0.67635E-
08*(T.^3))+(0.21556E-11*(T.^4));
% 1.987 is value of R(universal gas constant), and the unit is cal/mol K.
ni1=1;
y1=x1*ni1;
% for N2
x2=1.987*(0.36748E+01)+(-0.12082E-02*(T))+(0.23240E-05*(T.^2))+(-0.63218E-
09*(T.^3))+(-0.22577E-12*(T.^4));
ni2=3.773;
y2=x2*ni2;
cp= 4.184*(1/138.2)*(y1+y2); % for conversion to kJ/kg K we
multiply cp by 4.184
plot (T,cp)
hold on
text(T(end), cp(end), '\phi=0')
%______________________________________________________________________

%-----------------For phi=0.5, xb=0-----------------%


% for Fuel(Gasoline)
t= T./1000;
X= (-0.55313+ (181.62*t)+(-97.787*t.^2)+(20.402*t.^3)+(-0.03095./t.^2));
nif=0.04;
Y=X*nif;
12

% for O2
x1=1.987*(0.36256E+01)+(-0.18782E-02*(T))+(0.70555E-05*(T.^2))+(-0.67635E-
08*(T.^3))+(0.21556E-11*(T.^4));
ni1=1;
y1=x1*ni1;
% for N2
x2=1.987*(0.36748E+01)+(-0.12082E-02*(T))+(0.23240E-05*(T.^2))+(-0.63218E-
09*(T.^3))+(-0.22577E-12*(T.^4));
ni2=3.773;
y2=x2*ni2;
cp= 4.184*(1/142.76)*(Y+y1+y2);
plot (T,cp)
text(T(end), cp(end), '\phi=0.5')

%----------------For phi=0.5, xb=0.4------------------%


% for fuel
X= (-0.55313+ (181.62*t)+(-97.787*t.^2)+(20.402*t.^3)+(-0.03095./t.^2));
nif=0.024;
Y=X*nif;
% for O2
x1=1.987*(0.36256E+01)+(-0.18782E-02*(T))+(0.70555E-05*(T.^2))+(-0.67635E-
08*(T.^3))+(0.21556E-11*(T.^4));
ni1=0.800;
y1=x1*ni1;
% for N2
x2=1.987*(0.36748E+01)+(-0.12082E-02*(T))+(0.23240E-05*(T.^2))+(-0.63218E-
09*(T.^3))+(-0.22577E-12*(T.^4));
ni2=3.773;
y2=x2*ni2;
% for CO2
x3= 1.987*(0.24008E+01)+(0.87351E-02*(T))+(-0.66071E-05*(T.^2))+(0.20022E-
08*(T.^3))+(0.63274E-15*(T.^4));
ni3=0.128;
y3=x3*ni3;
% for H2O
x4= 1.987*(0.40701E+01)+(-0.11084E-02*(T))+(0.41521E-05*(T.^2))+(-
0.296371E-08*(T.^3))+(0.80702E-12*(T.^4));
ni4=0.144;
y4=x4*ni4;
cp= 4.184*(1/142.76)*(Y+y1+y2+y3+y4);
plot (T,cp,'--r')
13

%____________________________________________________________________

%----------------for phi=0.8, xb=0-----------------%


X= (-0.55313+ (181.62*t)+(-97.787*t.^2)+(20.402*t.^3)+(-0.03095./t.^2));
nif=0.064;
Y=X*nif;
% for O2
x1=1.987*(0.36256E+01)+(-0.18782E-02*(T))+(0.70555E-05*(T.^2))+(-0.67635E-
08*(T.^3))+(0.21556E-11*(T.^4));
ni1=1;
y1=x1*ni1;
% for N2
x2=1.987*(0.36748E+01)+(-0.12082E-02*(T))+(0.23240E-05*(T.^2))+(-0.63218E-
09*(T.^3))+(-0.22577E-12*(T.^4));
ni2=3.773;
y2=x2*ni2;
cp= 4.184*(1/145.49)*(Y+y1+y2);
plot (T,cp)
text(T(end), cp(end), '\phi=0.8')

%---------------for phi=0.8, xb=0.4----------------%


% for fuel
X= (-0.55313+ (181.62*t)+(-97.787*t.^2)+(20.402*t.^3)+(-0.03095./t.^2));
nif=0.0384;
Y=X*nif;
% for O2
x1=1.987*(0.36256E+01)+(-0.18782E-02*(T))+(0.70555E-05*(T.^2))+(-0.67635E-
08*(T.^3))+(0.21556E-11*(T.^4));
ni1=0.68;
y1=x1*ni1;
% for N2
x2=1.987*(0.36748E+01)+(-0.12082E-02*(T))+(0.23240E-05*(T.^2))+(-0.63218E-
09*(T.^3))+(-0.22577E-12*(T.^4));
ni2=3.773;
y2=x2*ni2;
% for CO2
x3= 1.987*(0.24008E+01)+(0.87351E-02*(T))+(-0.66071E-05*(T.^2))+(0.20022E-
08*(T.^3))+(0.63274E-15*(T.^4));
ni3=0.2048;
14

y3=x3*ni3;
% for H2O
x4= 1.987*(0.40701E+01)+(-0.11084E-02*(T))+(0.41521E-05*(T.^2))+(-
0.296371E-08*(T.^3))+(0.80702E-12*(T.^4));
ni4=0.2304;
y4=x4*ni4;
cp= 4.184*(1/145.496)*(Y+y1+y2+y3+y4);
plot (T,cp,'--r')

%__________________________________________________________________________
%--------------------for phi=1, xb=0----------------%
% for fuel
X= (-0.55313+ (181.62*t)+(-97.787*t.^2)+(20.402*t.^3)+(-0.03095./t.^2));
nif=0.08;
Y=X*nif;
% for O2
x1=1.987*(0.36256E+01)+(-0.18782E-02*(T))+(0.70555E-05*(T.^2))+(-0.67635E-
08*(T.^3))+(0.21556E-11*(T.^4));
ni1=1;
y1=x1*ni1;
% for N2
x2=1.987*(0.36748E+01)+(-0.12082E-02*(T))+(0.23240E-05*(T.^2))+(-0.63218E-
09*(T.^3))+(-0.22577E-12*(T.^4));
ni2=3.773;
y2=x2*ni2;
cp= 4.184*(1/147.32)*(Y+y1+y2);
plot (T,cp)
text(T(end), cp(end), '\phi=1')

%---------------for phi=1, xb=0.4-----------------%


% for fuel
X= (-0.55313+ (181.62*t)+(-97.787*t.^2)+(20.402*t.^3)+(-0.03095./t.^2));
nif=0.048;
Y=X*nif;
% for O2
x1=1.987*(0.36256E+01)+(-0.18782E-02*(T))+(0.70555E-05*(T.^2))+(-0.67635E-
08*(T.^3))+(0.21556E-11*(T.^4));
ni1=0.6;
y1=x1*ni1;
% for N2
15

x2=1.987*(0.36748E+01)+(-0.12082E-02*(T))+(0.23240E-05*(T.^2))+(-0.63218E-
09*(T.^3))+(-0.22577E-12*(T.^4));
ni2=3.773;
y2=x2*ni2;
% for CO2
x3= 1.987*(0.24008E+01)+(0.87351E-02*(T))+(-0.66071E-05*(T.^2))+(0.20022E-
08*(T.^3))+(0.63274E-15*(T.^4));
ni3=0.256;
y3=x3*ni3;
% for H2O
x4= 1.987*(0.40701E+01)+(-0.11084E-02*(T))+(0.41521E-05*(T.^2))+(-
0.296371E-08*(T.^3))+(0.80702E-12*(T.^4));
ni4=0.288;
y4=x4*ni4;
cp= 4.184*(1/147.32)*(Y+y1+y2+y3+y4);
plot (T,cp,'--k')

%____________________________________________________________________
%----------------for phi=1.2, xb=0---------------%
% for fuel
X= (-0.55313+ (181.62*t)+(-97.787*t.^2)+(20.402*t.^3)+(-0.03095./t.^2));
nif=0.096;
Y=X*nif;
% for O2
x1=1.987*(0.36256E+01)+(-0.18782E-02*(T))+(0.70555E-05*(T.^2))+(-0.67635E-
08*(T.^3))+(0.21556E-11*(T.^4));
ni1=1;
y1=x1*ni1;
% for N2
x2=1.987*(0.36748E+01)+(-0.12082E-02*(T))+(0.23240E-05*(T.^2))+(-0.63218E-
09*(T.^3))+(-0.22577E-12*(T.^4));
ni2=3.773;
y2=x2*ni2;
cp= 4.184*(1/149.144)*(Y+y1+y2);
plot (T,cp)
text(T(end), cp(end), '\phi=1.2')
%------------------for phi=1.2, xb=0.4-------------------------%
% for fuel
X= (-0.55313+ (181.62*t)+(-97.787*t.^2)+(20.402*t.^3)+(-0.03095./t.^2));
nif=0.0576;
Y=X*nif;
16

% for O2
x1=1.987*(0.36256E+01)+(-0.18782E-02*(T))+(0.70555E-05*(T.^2))+(-0.67635E-
08*(T.^3))+(0.21556E-11*(T.^4));
ni1=0.6;
y1=x1*ni1;
% for N2
x2=1.987*(0.36748E+01)+(-0.12082E-02*(T))+(0.23240E-05*(T.^2))+(-0.63218E-
09*(T.^3))+(-0.22577E-12*(T.^4));
ni2=3.773;
y2=x2*ni2;
% for CO2
x3= 1.987*(0.24008E+01)+(0.87351E-02*(T))+(-0.66071E-05*(T.^2))+(0.20022E-
08*(T.^3))+(0.63274E-15*(T.^4));
ni3=-0.0512;
y3=x3*ni3;
% for H2O
x4= 1.987*(0.40701E+01)+(-0.11084E-02*(T))+(0.41521E-05*(T.^2))+(-
0.296371E-08*(T.^3))+(0.80702E-12*(T.^4));
ni4=0.544;
y4=x4*ni4;
% for CO
x5= 1.987*(0.37101E+01)+(-0.16191E-02*(T))+(0.36924E-05*(T.^2))+(-0.20320E-
08*(T.^3))+(0.23953E-12*(T.^4));
ni5=0.3584;
y5=x5*ni5;
% for H2
x6= 1.987*(0.30574E+01)+(0.26765E-02*(T))+(-0.58099E-05*(T.^2))+(0.55210E-
08*(T.^3))+(-0.18123E-11*(T.^4));
ni6=-0.1984;
y6=x6*ni6;
cp= 4.184*(1/149.144)*(Y+y1+y2+y3+y4+y5+y6);
plot (T,cp,'--r')

%________________________________________________________________________
%-------------for phi=1.5, xb=0------------------------%
% for fuel
X= (-0.55313+ (181.62*t)+(-97.787*t.^2)+(20.402*t.^3)+(-0.03095./t.^2));
nif=0.12;
Y=X*nif;
% for O2
17

x1=1.987*(0.36256E+01)+(-0.18782E-02*(T))+(0.70555E-05*(T.^2))+(-0.67635E-
08*(T.^3))+(0.21556E-11*(T.^4));
ni1=1;
y1=x1*ni1;
% for N2
x2=1.987*(0.36748E+01)+(-0.12082E-02*(T))+(0.23240E-05*(T.^2))+(-0.63218E-
09*(T.^3))+(-0.22577E-12*(T.^4));
ni2=3.773;
y2=x2*ni2;
cp= 4.184*(1/151.88)*(Y+y1+y2);
plot (T,cp)
text(T(end), cp(end), '\phi=1.5')

%-------------for phi=1.5, xb=0.4--------------------------%


% for fuel
X= (-0.55313+ (181.62*t)+(-97.787*t.^2)+(20.402*t.^3)+(-0.03095./t.^2));
nif=0.072;
Y=X*nif;
% for O2
x1=1.987*(0.36256E+01)+(-0.18782E-02*(T))+(0.70555E-05*(T.^2))+(-0.67635E-
08*(T.^3))+(0.21556E-11*(T.^4));
ni1=0.6;
y1=x1*ni1;
% for N2
x2=1.987*(0.36748E+01)+(-0.12082E-02*(T))+(0.23240E-05*(T.^2))+(-0.63218E-
09*(T.^3))+(-0.22577E-12*(T.^4));
ni2=3.773;
y2=x2*ni2;
% for CO2
x3= 1.987*(0.24008E+01)+(0.87351E-02*(T))+(-0.66071E-05*(T.^2))+(0.20022E-
08*(T.^3))+(0.63274E-15*(T.^4));
ni3=0.1684;
y3=x3*ni3;
% for H2O
x4= 1.987*(0.40701E+01)+(-0.11084E-02*(T))+(0.41521E-05*(T.^2))+(-
0.296371E-08*(T.^3))+(0.80702E-12*(T.^4));
ni4=0.2476;
y4=x4*ni4;
% for CO
x5= 1.987*(0.37101E+01)+(-0.16191E-02*(T))+(0.36924E-05*(T.^2))+(-0.20320E-
08*(T.^3))+(0.23953E-12*(T.^4));
18

ni5=0.2156;
y5=x5*ni5;
% for H2
x6= 1.987*(0.30574E+01)+(0.26765E-02*(T))+(-0.58099E-05*(T.^2))+(0.55210E-
08*(T.^3))+(-0.18123E-11*(T.^4));
ni6=0.1844;
y6=x6*ni6;
cp= 4.184*(1/151.88)*(Y+y1+y2+y3+y4+y5+y6);
plot (T,cp,'--r')

%------------------------------------------------------------------------
xlabel ('Temperature (K)')
ylabel ('c_p_,_u (kJ/kg K)')
19

Output of Assignment1_Part2:

Fig no: 2: Specific heat at constant pressure of unburned gasoline, air, burned gas mixtures as
function of temperature, equivalence ratio, and burned gas fraction.

References:

[1] Internal Combustion Engine Fundamental, John B. Heywood.


1

Stoichiometry and Low Temperature Combustion Modeling

A stoichiometric reaction is defined such that the fuel burns completely and the only
products are carbon dioxide and water. The chemical equation can be represented as
C H O N . Now let us write the following stoichiometric reaction.

C H O N + (O25 3.76 N 2 ) n 1 CO 2 + n 2 H 2 O + n 1 N 2

The stoichiometric molar air- fuel ratio and the moles (i = 1, 2, 3) that describe the product
composition. Note that the stoichiometric reaction is expressed per mole of fuel. We know that
atoms are conserved, so we can write as

C: n1

H: 2n2

O: 2 2n1 n2

N: 2 * 3.76 * 2n

Solution of these four equations gives

/ 4 / 2

n1

n2 / 2

n3 / 2 3.76( / 4 / 2)

The stoichiometric air-fuel ratio A 1 is

A 1 = 28.85(4.76 ) /(12.01 1.008 16.00 14.01 )


2

Typical values are given by the table. Notice that for hydrocarbons ( 0) the stoichiometric
air fuel ratio by mass is not a particularly strong function of type . For typical hydrocarbon fuels
the stoichiometric air fuel ratio and the molar stoichiometric air fuel ratio and the product mole
fraction.

The mole and mass fraction of fuel in a stoichiometric fuel air mixture are as follows,
The fuel air equivalence ratio, is defined as the stoichiometric air fuel ratio, F divided by the
stoichiometric fuel air ratio, F 1 .

Mole fraction y 1 1 /(1 4.76 )

Mass fraction 1 /(1 As )

( A1 / A) ( F / Fs )

The equivalence ratio has the same value on a mole or mass basis. If 1 the mixture is called
lean, if 1 the mixture is said to be rich, and if 1 the mixture is said to be stoichiometric.
At low temperature (T < 1000K, such as in the exhaust) and carbon to oxygen ratios less than
one, the overall combustion reaction for any equivalence ratio can be written as,

C H O N + (O25 3.76 N 2 ) n 1 CO 2 + n 2 H 2 O + n 1 N 2

For reactant C/o ratios greater than one we would have to add solid carbon C(s) and
several other species to the product list, as we shall see. This equation assumes the dissociation

of molecules is negligible and a more general case is treated. For lean combustion ( 1)
products at low temperature, we will assume no product composition since there are four
equations and four unknowns.

For rich combustion we will assume that there is no product O 2 . In this case there are
five unknowns, so we need an additional equation to supplement the four atom balance equation.
We assume equilibrium consideration between the product species Co2, water and H2 determine
the product composition. This reaction is termed the water gas reaction:

CO2 + H 2 CO H 2 O
3

The Equilibrium constant K(T) equation is a curve fit of JANAF table data for 400 < T <
3200K :

ln K (T ) 2.743 (1.761 / t ) (1.611 / t 2 ) (0.2803 / t 3 )

Solutions for both rich and lean cases are given and the rich case, the parameter n is given
by the solution of the quadratic equation

n5 b1 ( b1 4ac) /( 2a)
2

** Mat lab Program 1

clc
clear all
%symbols (alpha,beta,gamma,delta);
alpha=4:1:12;
beta=10:2:26;
gamma=0;
delta=0;

A=alpha+(beta./4)-(gamma./2)
n1=alpha
n2=(beta./2)
n3=((delta./2)+3.76.*(alpha+(beta./4))-(gamma./2));
s=(n1+n2+n3);
mf1=(n1/s);
mf2=(n2/s);
mf3=(n3/s);

Output
4

A =

6.5000 8.0000 9.5000 11.0000 12.5000 14.0000


15.5000 17.0000 18.5000

n1 =

4 5 6 7 8 9 10 11 12

n2 =

5 6 7 8 9 10 11 12 13
** Mat lab Program 2

clc
close all
T=400:10:3200;
t = T./1000;
lnK=2.743-((1.761)./(t))-((1.611)./(t.^(2)))+((0.283)./(t.^(3)));
K=10.^(lnK);
figure(1)
plot(t,lnK,'-b')
xlabel('t');
ylabel('lnk')
grid on
figure(2)
plot(T,lnK,'-r')
xlabel('T');
ylabel('lnk')
grid on
figure(3)
plot(T,K,'-g')
xlabel('T')
ylabel('k')
grid on
5

Output

0
lnk

-2

-4

-6

-8
0 0.5 1 1.5 2 2.5 3 3.5
t

0
lnk

-2

-4

-6

-8
0 500 1000 1500 2000 2500 3000 3500
T
6

120

100

80

60
k

40

20

0
0 500 1000 1500 2000 2500 3000 3500
T
Specific enthalpy calculation for different alternative fuels

As different fuels have their different properties with respect to temperature and other parameters.

Talking about the internal energy of any fuel, it depends of few of the main parameters such as enthalpy
of the fuel, molar mass and temperature, where on other hand; enthalpy depends on the some constants
that are taken from the datum of Carbon, Hydrogen, Oxygen and Nitrogen at reference temperature of
273.15 K. Chemical analysis of the fuel was performed to obtain the H/C ratio, average molecular
weight, heating value, and the weight percent of aromatics, olefins, and total paraffin (including
cycloparaffins).

Starting from the basic equation, i.e. equation for enthalpy,


= + + + + + 8 8 . /

Where, Af1 to Af8 are the constants or coefficients for the molar enthalpy of different fuels, given by
Heywood, calculated through experiments, given below in the table

Fuel Af1 Af2 Af3 Af4 Af5 Af6 Af8 Molar


Mass
Methane -0.29149 26.327 -10.610 1.5656 0.1657 -18.331 4.300 16.04
Propane -1.4867 74.339 -39.065 8.0543 0.01219 -27.313 8.852 44.10
Hexane -20.777 210.48 -164.125 52.832 0.56635 -39.836 15.611 86.18
Isooctane -0.55313 181.62 -97.787 20.402 -0.03095 -60.751 20.232 11402
Methanol -2.7059 44.168 -27.501 7.2193 0.20299 -48.288 5.3375 32.04
Ethanol 6.99 39.741 -11.926 0 0 -60.214 7.6135 46.07
Regular -24.078 256.63 -201.68 64.750 0.5808 -27.561 17.792 114.8
Premium -22.501 228 -177.26 56.048 0.4845 -17.578 15.235 106.4
Diesel -9.1063 246.97 -143.74 32.329 0.0518 -50.128 23.514 148.6
In the formula of hf, incorporating all the values and finding values for all the fuels.
Now, molar internal energy is calculated from the below formula

Where, R is gas constant value equal to 8.314 KJ/Kg and T be the temperature which is variable and can
vary from 0K to any arbitrary value (we had taken3000K) up to which we want to measure the changes
take place in internal energy.

Now the specific internal energy is calculated from following formula

Where, Mf is the molar mass of the fuel.

Now when in programming in MATLAB is done by either in 2 ways


1) By separately making table for coefficients and corporating it by using commands.
2) By writing all the values itself in program.

I am using 2nd method as I found it convenient.


MATLAB Program

% Mf is molar mass
T = 0:10:3000;
v = T./1000;
R = 8.314;
% AfMe's are coefficients for Methane
Af1Me = -0.29141;
Af2Me = 26.327;
Af3Me = -10.61;
Af4Me = 1.5656;
Af5Me = 0.16573;
Af6Me = -18.331;
Af8Me = 4.3;
MfMe = 16.04;
hfMe = (Af1Me * v + (Af2Me.*v.^2)/2 + (Af3Me.*v.^3)/3 +
(Af4Me.*v.^4)/4 - (Af5Me./v) + Af6Me + Af8Me).*4186.6;
ufMe = hfMe - R.*T;
UfMe = ufMe./MfMe;
% Afpro's are coefficients for Propane
Af1Pro = -1.4867;
Af2Pro = 74.339;
Af3Pro = -39.065;
Af4Pro = 8.0543;
Af5Pro = 0.01219;
Af6Pro = -27.313;
Af8Pro = 8.852 ;
MfPro = 44.10;
hfPro = (Af1Pro.*v + ((1/2).*Af2Pro.*v.^2) + ((1/3).*Af3Pro.*v.^3) +
((1/4).*Af4Pro.*v.^4) - (Af5Pro./v) + Af6Pro + Af8Pro)*4186.6;
ufPro = hfPro - R.*T;
UfPro = ufPro./MfPro;
% AfHe's are coefficients for Hexane
Af1He = -20.777;
Af2He = 210.48;
Af3He = -164.125;
Af4He = 52.832;
Af5He = 0.56635;
Af6He = -39.836;
Af8He = 15.611;
MfHe = 86.18;
hfHe = (Af1He.*v + ((1/2).*Af2He.*v.^2) + ((1/3).*Af3He.*v.^3) +
((1/4).*Af4He.*v.^4) - (Af5He./v) + Af6He + Af8He)*4186.6;
ufHe = hfHe - R.*T;
UfHe = ufHe./MfHe;
% AfIsoc's are coefficients for Isooctane
Af1Isoc = -0.55313;
Af2Isoc = 181.62;
Af3Isoc = -97.787;
Af4Isoc = 20.402;
Af5Isoc = 0.03095;
Af6Isoc = -60.751;
Af8Isoc = 20.232;
MfIsoc = 114.2;
hfIsoc = (Af1Isoc.*v + ((1/2).*Af2Isoc.*v.^2) + ((1/3).*Af3Isoc.*v.^3)
+ ((1/4).*Af4Isoc.*v.^4) - (Af5Isoc./v) + Af6Isoc + Af8Isoc)*4186.6;
ufIsoc = hfIsoc - R.*T;
UfIsoc = ufIsoc./MfIsoc;
% AfMet's are coefficients for Methanol
Af1Met = -2.7059;
Af2Met = 44.168;
Af3Met = -27.501;
Af4Met = 7.2193;
Af5Met = 0.20299;
Af6Met = -48.288;
Af8Met = 5.3375;
MfMet = 32.04;
hfMet = (Af1Met.*v + ((1/2).*Af2Met.*v.^2) + ((1/3).*Af3Met.*v.^3) +
((1/4).*Af4Met.*v.^4) - (Af5Met./v) + Af6Met + Af8Met)*4186.6;
ufMet = hfMet - R.*T;
UfMet = ufMet./MfMet;
% AfEth's are coefficients for Ethanol
Af1Eth = 6.99;
Af2Eth = 39.741;
Af3Eth = -11.926;
Af4Eth = 0;
Af5Eth = 0;
Af6Eth = -60.214;
Af8Eth = 7.6135;
MfEth = 46.07;
hfEth = (Af1Eth.*v + ((1/2).*Af2Eth.*v.^2) + ((1/3).*Af3Eth.*v.^3) +
((1/4).*Af4Eth.*v.^4) - (Af5Eth./v) + Af6Eth + Af8Eth)*4186.6;
ufEth = hfEth - R.*T;
UfEth = ufEth./MfEth;
% AfReg's are coefficients for Regular
Af1Reg = -24.078;
Af2Reg = 256.63;
Af3Reg = -201.68;
Af4Reg = 64.750;
Af5Reg = 0.5808;
Af6Reg = -27.561;
Af8Reg = 17.792;
MfReg = 114.8;
hfReg =( Af1Reg.*v + ((1/2).*Af2Reg.*v.^2) + ((1/3).*Af3Reg.*v.^3) +
((1/4).*Af4Reg.*v.^4) - (Af5Reg./v) + Af6Reg + Af8Reg)*4186.6;
ufReg = hfReg - R.*T;
UfReg = ufReg./MfReg;
% AfPre's are coefficients for Premium
Af1Pre = -22.501;
Af2Pre = 228;
Af3Pre = -177.26;
Af4Pre = 56.048;
Af5Pre = 0.4845;
Af6Pre = -17.578;
Af8Pre = 15.235;
MfPre = 106.4;
hfPre =( Af1Pre.*v + ((1/2).*Af2Pre.*v.^2) + ((1/3).*Af3Pre.*v.^3) +
((1/4).*Af4Pre.*v.^4) - (Af5Pre./v) + Af6Pre + Af8Pre)*4186.6;
ufPre = hfPre - R.*T;
UfPre = ufPre./MfPre;
% AfDie's are coefficients for Diesel
Af1Die = -9.1063;
Af2Die = 246.97;
Af3Die = -143.74;
Af4Die = 32.329;
Af5Die = 0.0518;
Af6Die = -50.128;
Af8Die = 23.514;
MfDie = 148.6;
hfDie = (Af1Die.*v + ((1/2).*Af2Die.*v.^2) + ((1/3).*Af3Die.*v.^3) +
((1/4).*Af4Die.*v.^4) - (Af5Die./v) + Af6Die + Af8Die)*4186.6;
ufDie = hfDie - R.*T;
UfDie = ufDie./MfDie;
plot (T, UfMe, 'r')
hold on
plot (T, UfPro, 'b');
hold on
plot (T, UfHe, 'c');
hold on
plot (T, UfIsoc, 'g');
hold on
plot (T, UfMet, 'm');
hold on
plot (T, UfEth, 'y');
hold on
plot (T, UfReg, 'k');
hold on
plot (T, UfPre, '-.r');
hold on
plot (T, UfDie, '-.g');
hold off
xlabel('Temp in K'), ylabel('Sp. Int Energy (kJ/kg)'), grid on
title('Internal Energy for Different fuels')

Plot obtained
Fig. Temperature vs. Internal Energy

Fig. Notation for different fuels.

Conclusions

Following things we conclude from the program and graph


1) With increase in temperature. There will be increase in the value of internal energy as both hf
and uf are dependent of temperature.
2) There will be some negative start will be there for all the fuels as the value of enthalpy depends
on Af6 andAf8 (JANAF coefficients) and when temperature tends to zero molar internal energy
equates with enthalpy only.
3) Higher the molar mass higher will be energy and lower the molar mass, lower will be the energy.
Thus, internal energy strongly affected by molar mass of fuel.

References
1) J. B. Heywood for Internal Combustion Engines Fundamentals.
Temperature and air-fuel ratio dependent specific heat ratio functions for
lean burned and unburned mixture

The most important thermodynamic property used in heat release calculations for engines is
the specific heat ratio. The functions proposed in the literature for the specific heat ratio are
temperature dependent and apply at or near stoichiometric air-fuel ratios. However, the specific
heat ratio is also influenced by the gas composition in the engine cylinder and especially becomes
important for lean combustion engines.

The most important thermodynamic property used in the heat release calculations for engines is
the specific heat ratio. Gatowski et al. developed a single zone heat release model based on the
first law of thermodynamics that has been widely used, where the specific heat ratio is
represented by a linear function of the mean charge temperature. Brunt et al. utilized a second
order function, evaluated for a spark ignition (SI) engine fuelled with C8H16. Egnell evaluated an
exponential specific heat ratio function and emphasized that the constants have to be chosen
carefully with respect to the influence of temperature and gas composition.

Specific heat ratio functions and importance of composition effects

The specific heats are dependent on charge temperature and composition and, as such, will
vary during the engine cycle and with operating conditions. Although the use of a temperature
dependent specific heat ratio function greatly reduces the potential errors in calculated the heat
release, the dependency of the specific heat on composition is greater at the lean side of
stoichiometric conditions.

Derived specific heat ratio functions

1. Function of Gatowski et al.

0 K1 (T Tref )

1000

where 0 is a reference value (1.38), K1 is a constant (0.08) and Tref is a reference temperature

1
(300 K).
Programme for Gatowski Function

function gatowski()
gamma0=1.38; %%%%%%%%%%%%%Reference value of Gamma
K1=0.08; %%%%%%%%%%%%%Constant
Tref=300; %%%%%%%%%%%%%Reference Temperature
T=[300:1:2500]; %%%%%%%%%%%%%Temperature range
gamma=gamma0-(K1*(T-Tref)/1000); %%%%%%%%%%%%%Gatowski Function Formula
figure(1)
plot(T, gamma)
xlabel Temperature[K];
ylabel Specific-heat-ratio;
grid ontext(1200,1.31,'Gatowski Function')

1.38

1.36

1.34

1.32
Specific-heat-ratio

Gatowski Function
1.3

1.28

1.26

1.24

1.22

1.2
0 500 1000 1500 2000 2500
Temperature[K]

Fig.1 Gatowski function for specific heat ratio

2
2. Function of Brunt et al.

1.338 6.0 10 5 T 1.0 10 8 T 2

where T is temperature in K.

function brunt()
T=[300:1:2500]; %%%%%%%%%%%%%Temperature range
gamma1=(1.338-(6.0*(10e-5).*T)+(1.0*(10e-8).*T.^2)); %%%%%%%%%%%%%Brunt
Function Formula
hold on
figure (1)
plot(T, gamma1,'r')
xlabel Temperature[K];
ylabel Specific-heat-ratio;
grid ontext(1200,0.78,'Brunt et al. Function')

1.4

1.3 Gatowski Function

1.2

1.1
Specific-heat-ratio

0.9

0.8
Brunt et al. Function

0.7

0.6

0.5

0.4
0 500 1000 1500 2000 2500
Temperature[K]

Fig. 2 Brunt function for specific heat ratio

3
3. Function of Engell

k2
( )
0 k1e T

where 0 is a reference value (1.38), k1 and k 2 are constants (0.2, 900) and Tref is a reference
temperature (300 K).

T=[300:1:2500];
gamma0=1.38;
k1=0.2;
k2=900;
gamma2=gamma0-k1*exp(-k2./T);
%hold on
figure (1)
plot(T, gamma2,'g')
xlabel Temperature[K];
ylabel Specific-heat-ratio;
grid on
text(400,1.36,'Engell Function')

1.38

1.36 Engell Function

1.34
Specific-heat-ratio

1.32

1.3

1.28

1.26

1.24
0 500 1000 1500 2000 2500
Temperature[K]

Fig. 3 Engellt function for specific heat ratio

4
4. Derived functions of burned and unburned mixtures

The air-fuel ratio was increased from 1.0 to 1.6 in steps of 0.1, and the
ranges of temperature were 300-1500 K and 300-2500 K for the unburned and burned mixtures,
respectively. The data of the specific heat ratio was used to derive functions for the unburned and
burned mixtures using a least squares approach and choosing the parameters as temperature (T in K)
and air-fuel ratio ( ). The functions of the specific heat ratio for the unburned and burned
mixtures in T and can be written as:
a7
u a1 a2T a3T 2 a4T 3 a5T 4 a6T 5

b3 b5 b6T b8 b9T b10T 2


b b1 b2T b4T
4
b7T
3

2 3 2

The values of the coefficients of these functions are given in Table 1.

Coefficients for specific heat ratio function for unburned and burned mixtures

Coefficients ( u ) Values Coefficients ( b ) Values

a1 1.464202464 b1 1.498119965

a2 0.000150666 b2 0.00011303

a3 7.34852e-08 b3 0.26688898

a4 1.55726E-10 b4 4.03642e-08

a5 7.6951E-14 b5 0.273428364

a6 1.19535E-17 b6 5.7462e-05

a7 0.063115275 b7 7.2026e-12

- b8 0.08218813

- b9 1.3029e-05

- - b10 2.35732e-08

5
function gammaub()
T=[300:1:2500];
a1=1.464202464;
a2=-0.000150666;
a3=-7.34852*10^-8;
a4=1.55726*10^-10;
a5=-7.6951*10^-14;
a6=1.19535*10^-17;
a7=-0.063115275;
for lambda=1.0:0.1:1.6;
gammau=((a1)+(a2.*T)+(a3.*T.^2)+(a4.*T.^3)+(a5.*T.^4)+(a6.*T.^5)+(a7./lambda)
);
plot(T, gammau,'r')
hold on
xlabel Temperature[K];
ylabel Specific-heat-ratio;
end
grid on
text(1100,1.29,'Gamma unburned Function')

1.4

1.35
Specific-heat-ratio

1.3
Gamma unburned Function

1.25

1.2

1.15
0 500 1000 1500 2000 2500
Temperature[K]

Fig. 4. Variation of specific heat ratio with temperature and air-fuel ratio for unburned mixture.

function gammab()
T=300:1:2500;
b1=1.498119965;
b2=-0.00011303;
b3=-0.26688898;
6
b4=4.03642e-08; _
b5=0.273428364;
b6=5.7462e-05;
b7=-7.2026e-12;
b8=-0.08218813;
b9=-1.3029e-05;
b10=2.35732e-08;
for lambda=1.0:0.1:1.6;
gammab=b1+(b2.*T)+(b3/lambda)%+(b4*(T.^2))+(b5/(lambda^2))+((b6.*T)/lambda)+(b7.*(T.^
3))+(b8/(lambda^3))+((b9.*T)/(lambda^2))+((b10.*(T.^2))/lambda));
plot(T,gammab,'b')
hold on
xlabel Temperature[K];
ylabel Specific-heat-ratio;
box on
grid on
end

1.3

1.25

1.2
Specific-heat-ratio

1.15

1.1

1.05

0.95

0.9
0 500 1000 1500 2000 2500
Temperature[K]

Fig. 5 Variation of specific heat ratio with temperature and air-fuel ratio for burned mixture.

7
1.4
Engell Function
Gatowski Function
1.3

1.2 Gamma unburned Functi

1.1
Specific-heat-ratio

0.9

0.8
Brunt et al. Function

0.7

0.6

0.5

0.4
0 500 1000 1500 2000 2500
Temperature[K]

Fig.6 Comparison between several unburned mixture specific heat ratio functions.

Procedure to run Matlab Function

To run functions in matlab load all function files.


Type function name in command window.
All the plots are Specific heat ratio against Temperature.
Standard temperature is given input in each function as 300 to 2500 K.
Functions generated are
Gatowski
Brunt
Engell
Comparision
Gammab
Gammaub

8
References

1. Klein M, Erikson L. A specific heat ratio model for single-zone heat release models.
SAE Paper No: 2004-01-1464, 2004.
2. Ferguson CR. Internal combustion engines-applied thermosciences. John Wiley & Sons;
1986.
3. Heywood JB. Internal combustion engine fundamentals. McGraw-Hill Inc.; 1988.

9
Narayan Susange

121495019

VARIOUS P-V DIAGRAM OF OTTO CYCLE


MATLAB PROGRAM FOR VARIABLE CR
To run this program , type VARIABLECR(V, P ) provide compression ratio in programm
or compression ratio values.

function pdot = VARIABLECR(V, P )


clear all
clc
Qin = 600; r = 1.4;
P=0.95;
CR =04;
T=339; Cp = 1.004; Cv = 0.717; R = 0.287; m = 1;
V = m.*R.*T./(P.*100);
T1 = linspace(T,T,100);
P1 = linspace(P,P,100);
V2 = linspace(V,V/CR,100);
P2 = P1.*(V./V2).^r;
V22 = V/CR;
P22 = P.*(CR).^r;
T2 = P2.*V2.*T1./(P1.*V);
T22 = P22.*V./CR.*T./(P.*V);
V3 = V/CR;
T3 = T2+(Qin./Cv);
T33 = T22+(Qin./Cv);
P3 = (P2.*T3)./T2;
P33 = (P22.*T33)./T22;
P_3= linspace(P22,P33,100);
V4 = linspace(V/CR,V,100);
P4 = P33.*(V3./V4).^r;
P44 = P33.*(1./CR).^r;
T4 = P4.*V4.*T3./(P3.*V3);
T44 = P44.*V.*T33./(P33.*V./6);
Qout = Cv.*(T44 - T);
OTTOEff = ((Qin - Qout)./Qin).*100;
WD = Qin - Qout;
MEP = WD./(V-(V/CR))./100;
P5 = linspace(P44,P,100);
figure(1);
plot(V,P,'-y',V2,P2,'-m',V3,P_3,'-b',V4,P4,'-c',V,P5,'-r')
box on % turn on the axes box
axis([0 1.2 0 60])
xlabel('VOLUME IN m^3') % label horizontal axis
ylabel('PRESSURE IN Bar ') % label vertical axis
title('P - V DIAGRAM FOR VSARIABLE CR OF OTTO CYCLE')
gtext ('___________CR = 04 ')
hold on;
Fig:- Pressure Vs Volume graph for varying Compression Ratio.
MATLAB PROGRAM FOR VARIABLE INPUT PRESSURE

To run this program , type VARIABLEINPUTPRESSUREPLOT(V, P ) provide pressure values


in programm or input pressure values.

function pdot = VARIABLEINPUTPRESSUREPLOT(V, P )


clear all
clc
Qin = 600; r = 1.4;
% P =input('Value of inlet pressure = ');
P = 0.5;
T=339; CR = 6; Cp = 1.004; Cv = 0.717; R = 0.287; m = 1;
V = 1;
T1 = linspace(T,T,100);
P1 = linspace(P,P,100);
V2 = linspace(V,V/6,100);
P2 = P1.*(V./V2).^r;
V22 = V/6;
P22 = P.*(6).^r;
T2 = P2.*V2.*T1./(P1.*V);
T22 = P22.*V./6.*T./(P.*V);
V3 = V/6;
T3 = T2+(Qin./Cv);
T33 = T22+(Qin./Cv);
P3 = (P2.*T3)./T2;
P33 = (P22.*T33)./T22;
P_3= linspace(P22,P33,100);
V4 = linspace(V/6,V,100);
P4 = P33.*(V3./V4).^r;
P44 = P33.*(1./CR).^r;
T4 = P4.*V4.*T3./(P3.*V3);
T44 = P44.*V.*T33./(P33.*V./6);
Qout = Cv.*(T44 - T);
OTTOEff = ((Qin - Qout)./Qin).*100;
WD = Qin - Qout;
MEP = WD./(V-(V/6))./100;
P5 = linspace(P44,P,100);
figure(1);
plot(V,P,V2,P2,V3,P_3,V4,P4,V,P5)
axis([0 1.2 0 60])
% plot(V,P,'-y',V2,P2,'-m',V3,P_3,'-b',V4,P4,'-c',V,P5,'-r')
box on % turn on the axes box
xlabel('VOLUME IN m^3') % label horizontal axis
ylabel('PRESSURE IN Bar ') % label vertical axis
title('P - V DIAGRAM FOR VARYING INPUT PRESSURE OTTO CYCLE')
text(V22,P33-P33+3,'V_c = 0.1667 ')
gtext ('___________pressure = 0.5 ')
hold on;
Fig:- Pressure Vs Volume graph for varying Input Pressure
MATLAB PROGRAM FOR VARIABLE INPUT PRESSURE
bore = input('Value of diameter of Engine Cylinder = ');
stroke = input('Value of piston stroke = ');
rc = input('Designated compression ratio = ');
Vs = (pi/4)*(bore^2)*stroke;
l = input('length of connecting rod = ');
R = 2*l/stroke;
thetas = -pi/9;
thetad = 4*pi/9;
theta = -pi:0.1:pi;
bb = ones(1,63);
i = 1 ;
while i < 64
if theta(:,i)< thetas
xb = 0;
elseif theta(:,i) > thetas + thetad
xb = 1;
else
xb = 1 - exp(-5*(theta(:,i)-thetas/thetad)^3);
end
bb(:,i) = xb;
i = i + 1;
end
subplot(2,2,1), plot(theta,bb)
xlabel('Crank Angle')
ylabel('Burn Fraction Xb')
axis([-pi pi 0 1])
thetad = 4*pi/9;
theta = -pi:0.1:pi;
bb = ones(1,63);
cc = ones(1,63);
i = 1 ;
while i < 64
if theta(:,i) < thetas
xb = 0;
elseif theta(:,i) > thetas + thetad
xb = 1;
else
xb = 1 - exp(-5*(theta(:,i)-thetas/thetad)^3);
end
bb(:,i) = xb;
i = i + 1;
end
i = 1;
while i < 64
if theta(:,i)< thetas
hr = 0;
elseif theta(:,i) > thetas + thetad
hr = 0;
else
hr = (5*3*1.8/thetad)*(1-bb(:,i))*(theta(:,i)-thetas/thetad)^2 ;
end
cc(:,i) = hr;
i = i + 1;
end
subplot(2,2,3),plot(theta,cc)
xlabel('Crank Angle (Radian)')
ylabel('Rate of heat release (MJ/Kg)')
axis('equal')
Vs = 7.8540e-004;
R = 3;
rc = 10;
thetas= -4*pi/18;
pdot =(-(6.*exp(-5.*((theta-thetas)/4.*pi/9).^3).*((theta-
thetas)/4.*pi/9).^2)/((Vs/rc-1)+(Vs/2).*((R+1-cos(theta))-(R.^2-
sin(theta).^2).^0.5)) +(0.7.*p.*Vs.*sin(theta).*(1+(cos(theta).*(R.^2 -
sin(theta).^2).^-0.5)))./((Vs/rc-1)+(Vs/2).*((R+1-cos(theta))-(R.^2 -
sin(theta).^2).^0.5)));
fig:- Burn friction and Rate of heat release Vs Crank Angle(Radian)

References:-
1. "Internal combustion engine applied thermo sciences" by Fergusson .
1

MODELING OF ENGINE IN CYLINDER THEMODYNAMICS


UNDER DIFFERENT VALUES OF IGNITION RETARD

Introduction

The spark timing plays a significant role in the combustion process and in deciding the
engine parameters of SI engine. This paper aims at demonstrating the effect of advanced and
retarded spark timing on the burn fraction variation versus crank angle, cumulative heat
release rate versus the crank angle and the pressure variation as a function of crank angle with
the help of MATLAB programs. For this purpose, a basic finite heat release model is used for
the combustion process in SI engines. This model can also be extended to evaluate effect of
spark timing on engine work and thermal efficiency. In various section of the paper, the
MATLAB codes used for analysis are provided for future research work. Salient results, such
as peak PRESSURE RATE crank angles for different spark timings , are derived from
analysis. The combustion is a two-zone burnt/unburned model with the fuel burning rate
described by a Wiebe function. Under extreme spark timing retard conditions, the Wiebe
function describing the heat release of the fuelair reactions was modified to account for the
critical change in pressure distribution in the cylinder due to the abnormal spark retard. An
empirical correlation for cylinder pressure variation during the mass blowdown process,
which occurs between the exhaust valve opened and bottom-dead-centre, was included in the
simulation to enhance the predictive capability of the engine model
2
3

1. ENGINE MODEL

A zero-dimensional thermodynamic cycle simulation with two-zone burnt/unburned


combustion model, mainly based on Fergusons [4] work, was developed to predict the
cylinder pressure, burnt and unburned temperatures of gases, work done, heat loss, enthalpy
of exhaust gas, mass flow rate, etc. The zero-dimensional model is based on the first law of
thermodynamics in which an empirical relationship between the fuel burning rate and crank
angle position is established. The two-zone combustion model assumes that at any instant of
time during the combustion, the cylinder volume is divided into burnt and unburned zones by
an infinitesimally thin flame front. Each zone is assumed to be in thermodynamic equilibrium
and both zones having the same uniform pressure at any instant of time. The region in the
combustion chamber is treated as a control volume as illustrated in figure 1. The governing
equations are the mass and energy conservation equations and equations of state. These
equations, with crank angle as the independent variable, form the building block of
this thermodynamic model.

Mass and energy balances

For a control volume enclosing the fuelair mixture, the rate of change of total mass of an
open system is equal to the sum of mass flowing into and out of the system:
Applying the first law of thermodynamics to an open thermodynamic system, the energy
equation is
The major assumptions and approximations used in the development include the
following:

(1) The thermodynamic system is the cylinder contents which are assumed to be spatially
homogeneous, of constant composition, and to obey the ideal gas equation of state;
(2) the specific heat capacities (and the ratio of specific heat capacities) are constant;
(3) most of the other thermodynamic properties (including pressure and temperature) vary
only with time (crank angle) and are spatially uniform;
(4) the flow rates are determined from quasi-steady, one-dimensional flow equations;
(5) the intake and exhaust manifolds are assumed to be infinite plenums containing gases
at constant temperature and pressure;
(6) the fuel is assumed to be completely vaporized and mixed with the incoming air;
(7) the combustion efficiency was assumed to be 100% (i.e. no unburnt fuel); and
(8) the blow-by was assumed to be zero.
4

Fig. 1 is a schematic of the engine cylinder which shows cylinder heat transfer, work, intake
flows, exhaust flows, and combustion. The first law of thermodynamics on a differential
rate basis for this systemis

(1)

where m is the total mass in the cylinder, u Qnet


is thenet energy (heat) input (energy) release from combustion* plus energy gain due to heat
transferto the cylinder gas), p is the cylinder pressure, V is the cylinder volume, m in is the
total massflow into the system, hin is the specific enthalpy into the system, m out is the total
mass flow outof the system, and hout is the specific enthalpy out of the system.

* The use of the energy (heat) input due to combustion is predicated on the use of enthalpy
values whichonly consist of the sensible component and do not contain the chemical
component.

Fig. 1.Schematic of the thermodynamic system.

Since the internal energy is only a function of temperature(i.e. no dependence on pressure or


composition), this allows equation (1) to be rewritten as

(2)

To solve equation (2), a number of items are needed. These include the total net energy(heat)
input, the cylinder volume change, and the mass flow rates. The rate of total energy(heat)
inputis given by
5

..(3)
where(dQ/d)app is the apparent heat release from combustion, and (dQ/d)heat is the
energygain to the cylinder gases due to heat transfer. The apparent heat release may be
determinedfrom an expression for the burned fuel mass. A number of such expressions exist;
for thiswork the sub-model chosen for the burned fuel mass fraction was the cosine law.
The derivative of this burn rate model yields the following apparent heat release rate due
tocombustion

(4)
whereLHV is the lower heating value of the fuel, m is the total charge mass, ois the start
ofcombustion, bis the combustion (burn) duration, and AF is the mass air to fuel ratio.

In general, the combustion duration, b, is a complex function of turbulence,


composition,chemistry, and other factors. Although the turbulence is known to increase with
engine speed,the increase is not thought to be enough to completely compensate for the
shorter availabletimes at the higher engine speeds. For this work, the combustion duration is
approximated by abasic function of engine speed and equivalence ratio. This relationship was
recommended byCampbell based on experimental data reported by Taylor. The relationship
is

(5)
whereN is the engine speed (rpm) and
enginespeed included assists in accounting for the greater number of crank angles involved in
thecombustion process as engine speed increases.

The heat transfer to the cylinder gasesis given by

(6)

wherehconv is the convective heat transfer coefficient, A() is the total surface area in
thecylinder subject to the convective heat transfer, Twall is the cylinder wall temperature, and
T isthe instantaneous average cylinder gas temperature. The basic convective heat transfer
coefficientused in this work is from Eichelberg, but others would be just as suitable. The total
surface area consists of the cylinder walls, the piston top, and the bottom of the cylinderhead,
and varies with crank angle. For this work, the cylinder wall temperature is assumed tobe
450K.
6

Thermal properties
In a two-zone burnt/unburned model, the unburned mixture and burnt mixture zones are
each treated as sep-arate open systems. Hence, the specific internal energy and volume are
given by

xu b 1 x uu . 7
U
u
m

xvb 1 x vu .................................................................(8)
V
v
m

Assuming that the pressures of burnt and unburned gases are equal,vb and vu are functions
of bothTb, Tu andp. Hence,

Fuel burning rate correlation


The fuel burning rate in the SI engine is normally modelled by the Wiebe function , however,
it is mod-ified in this study to include the effect of HVIR on theheat release pattern. The
burnt mass fraction is given by

ig
m 1

x( ) 1 exp a a5 m 1.5 ..............(9)
b

where the efficiency parameter is a=5 and the shape factor is m=1:5 for the engine used in
this study.

Trapped mass in control volume


The trapped mass in the control volume at various periods is given by

V ( )
m when IVC 360 0 CA
vu

V ( )
m when EVO 360 0 CA
vb
7

The cylinder volume at any crank angle is given by

r 1
V ( ) VC 1
1

0.5
1 cos 1 ( sin ) (10)
2
2 2

s
=
2l

Mass blowdown
In addition to the above processes in the previous sec-tions, the mass blowdown process,
which occurs between is also considered in this model to enhance the predictive capability of
the Modelling of engine in-cylinder thermodynamics under high values of ignition retard
model. An empirical correlation for the blowdown pres-sure is obtained as in the following
form.

EVO
p( ) p EVO exp a a 5 (11)

180 EVO
0

MATLAB PROGRAMS:

1.For Fuel burning rate


INPUT: a=5.................. efficiency parameter which is function of load
m=1.5...............shape factor
theta0 = -15spark timing (CA)
dtheta= 80combustion duration( CA)

Program:

function burnt_frac = xb(theta0,dtheta)

theta = [-15:1:60];

theta0 = -15 ;

for dtheta= 60:5:80 ;

a=5 ;
m=2 ;
burnt_frac=1-exp(-a*(((theta-theta0)./(dtheta)).^(m+1))) ;
plot(theta,burnt_frac);
title('Burnt mass fraction variation')
8

xlabel('theta');
ylabel('burnt mass fraction');
hold on
grid on

text(0.6,0.98,'80 deg');

end

GRAPH

2. Trapped mass in control volume

Input:

Bore diameter= 80 mm
Compression ratio=6
Length of connecting rod=130 mm
Unburnt specific valume=0.125 m3/kg
Burnt specific volume=0.485 m3/kg

function mass_controll=trapped_mass(b,r,l,Vu,Vb)
9

b=0.08; %input('enter the value of bore diameter=')%


Ah=(pi*b^2)/2;
Ap=Ah/2;
r=6; %input('enter the value of compression ratio=');
s=0.25; %input('enter the value of stroke lenght=');
l=0.13; %input('enter the lenght of connecting rod =');
q=s/(2*l);
Vswept=Ap*s;
Vc=Vswept/(r-1);
theta=-360:360;
Vtheta=Vc.*(1+((r-1)/2).*(1-cosd(theta)+(1/q).*(1-(1-
q.^2.*(sind(theta)).^2).^0.5)));
Vu=0.125;
Vb=0.485;
if (-360<theta<=-180);
mass=Vtheta./Vu;
else
mass=Vtheta./Vb;
end
plot(theta,mass)
title('trapped mass in controll volume');
xlabel('theta(degg)');
ylabel('mass trapped(kg)')
end

GRAPH

3.For pressure rate with respect to theta.

Input:

Calorific value=42000 kj/kg


10

Mass=1 kg
Rpm=5000
Equivalence ratio=0.9
Air fuel ratio=14.67
Spark timing=20 deg before BDC
Compression ratio=9.5
Initial pressure=101.3 kpa
Initial temperature=300 k
Exust pressure= 500 kpa

Program
function dpdtvariation=pressurerate(LHV,m,N,phi,AF,thetaO,r,P1,T1,P4)
clear all
clc
LHV = 42000;
m = 1;
N = 5000;
phi = 0.9;
AF = 14.67;

thetab = 40 + 5 * ((N/600)-1) + 166 * ((phi - 1.1)^2);


fprintf('For Combustion duration (degrees): %f\n',thetab);
thetaB = pi*thetab/180;
thetaO=340;

thetaO = thetaO*pi/180; %ignition Delay = 20 degrees


theta = linspace(0, 4*pi, 100);

HR_app = (LHV .* m .* pi * sin(pi.*(theta - thetaO)/thetab))./(2 .* thetab


.* (AF + 1));

hconv = 40;
A = 2 * pi * (0.092)^2 + pi * 0.092 * sin(theta) * 0.084;
Twall = 450;

R = 0.287; %Defining Constants, for Otto


Cp = 1.004;
Cv = 0.743;
gm = Cp/Cv;
r = 9.5;
Qout = 200;

P1 =101.3 %input('Intake Pressure: '); %Values for point 1%


T1 =300 %input('Intake Temperature: ');%
V1 = (m*R*T1)/P1;

P2 = P1*(r^gm); %Values for point 2


T2 = T1*(r^(gm-1));
V2 = V1/r;

P4 =500; %input('Exhaust Pressure: '); %Values for point 4%


T4 = (T1*P4)/P1;
V4 = V1;

V3 = V2; %Values for point 3


T3 = T4 * (r^(gm-1));
11

P3 = P2 * (T3/T2);

Temp1 = linspace(T1, T2, 25);


Temp2 = linspace(T2, T3, 25);
Temp3 = linspace(T3, T4, 25);
Temp4 = linspace(T4, T1, 25);
T = [Temp1, Temp2, Temp3, Temp4];

HR_heat = hconv .* A .* (T - Twall);

HR_net = HR_app + HR_heat;

tht = linspace(0,720,100);

p = 1;
V = 3.35*10^-3;
Vc = 0.005 * V;
Vtheta = Vc + V .* cos(theta);

% dvdt = Vtheta ./ theta;


dvdt(1)= Vc;
for n=1:99
dvdt(n+1) = (Vtheta(n+1) - Vtheta(n)) / (theta(n+1) - theta(n));
end;

m_in = 30; %Assuming Mass flow rates


m_out = 20;
h_in = 300;
h_out = 420;
Prate = ((gm - 1)/V) * (HR_net - (gm/(gm-1)) .* p .* dvdt + m_in * h_in -
m_out * h_out);

plot(tht, Prate);
xlabel('Crank Angle');
ylabel('Pressure Rate');
title('Pressure Rate vs Crank Angle');
end
12

Graph:

4.Blowdown pressure

Pevo=2.8;
Patm=1.0135;
theta_evo=129.*(pi/180);
A=-log(Patm/Pevo);
m=0.45;
theta=-360:360;
theta=theta.*(pi/180);
P_theta=Pevo.*exp(-A.*((theta-theta_evo)./(180-theta_evo)).^(m+1));
theta=theta.*(180/pi);
plot(theta,P_theta);
title('Blowdown pressure');
xlabel('theta');
ylabel('blow down pressure');
grid on
13

Graph:

References

1.Internal combustion engines by Heywood

2. Illustration of the Use of an Instructional Version of a Thermodynamic Cycle


Simulation for a Commercial Automotive Spark-Ignition Engine by JERALD A.
CATON, Texas A&M University

3.Annand WJ (1963) Heat Transfer in Cylinder of Reciprocating Internal Combustion


Engines. Proc Inst Mech Engineers, vol 177, No 36, pp 973-990
14
Copyrighted Material

Zero Dimensional Single Zone Modeling of SI Engine

Prof. D. N. Malkhede
Professor in Mechanical Engineering,
Government College of Engineering, Pune
Email: dnm.mech@coep.ac.in

The heat release rate describes the temporal course of energy release in the combustion chamber.
In order to model combustion in the context of process calculation we use various approaches
and mathematical modelings, which all have the goal of describing the actual heat release via
combustion as exactly as possible as so-called substitute heat release rates. Wiebe/Vibe function
is one of the substitute methods to describe the heat release rate.

1. INTRODUCTION

The combustion chamber of an internal combustion engine is limited by the combustion chamber
walls, the piston, and the valves. The combustion chamber walls simultaneously represent the
system boundaries.

Fig.1- Single-zone cylinder model

The basis for the majority of the heat-release models is the first law of thermodynamics & the
energy conservation equation.

The energy balance or the 1st law of thermodynamics provides for the cylinder neglecting the
kinetic energy

Where,
E= energy [ J ]
h=specific enthalpy [J / kg]
Q =heat release [kJ / KW]
m=mass [kg]

Some commonly made assumptions for the single-zone models are


a) the cylinder contents and the state is uniform throughout the entire chamber
b) the combustion is modeled as a release of heat
c) the heat released from the combustion occurs uniformly in the chamber
d) the gas mixture is an ideal gas
Copyrighted Material

Heat Release Rate by Wiebe / Vibe Function

Prof. D. N. Malkhede
Professor in Mechanical Engineering,
Government College of Engineering, Pune
Email: dnm.mech@coep.ac.in

THE VIBE SUBSTITUTE HEAT RELEASE RATE

Proceeding from the "triangular combustion", Vibe (1970) provided the following relation for
the total heat release rate with the help of reaction kinetic considerations.

The entire amount of released energy is thereby to be calculated from the product of the fuel
mass brought into the combustion chamber and the lower heating value.
Furthermore, SOC signifies the start of combustion and CD the combustion duration. The so-
called Vibe form parameter is designated with m.

At the end of combustion, i.e. at = EOC, a certain percentage conv,total of the total energy
added with the fuel must be converted, for which is valid

From this follows for the factor a the relation

a = ln(1 conv,total )

from which we obtain the following numerical values

If we derive the total heat release rate from the degree of crank angle, we obtain expression for
the heat release rate.

Fig. 2 shows heat release rates for various Vibe form factors. One should note here that Vibe
form factors smaller than 0 are also possible. In order to adjust a real combustion by means of a
Vibe substitute heat release rate, there are various methods for determining the three Vibe
parameters start of combustion, combustion duration, and form parameter. The parameters can
either be adjusted visually or can be determined using mathematical methods (e.g. that of the
least square). It is however important that important process data such as peak pressure,
indicated mean pressure and the exhaust gas temperature are calculated with the substitute heat
release rate in agreement with reality. The determination of the Vibe parameters is solidly
integrated into most pressure analysis programs.

Fig.2 The Vibe substitute heat release rate for different Vibe shape coefficients

2.1 Matlab Script cal_a.m

Conveff = 90 : 0.2 : 100 ;


a = -log(1-(Conveff/100));
plot(Conveff, a, 'o-')
xlabel('Conversion Efficiency in %')
ylabel('Parameter a')
title('Parameter a Versus Conversion Efficiency')
grid
2.2. Function description and Matlab Program

2.2. Program Code-Total Combustion Progression


function Tothrr=calhrr1(Conveff,m,theta1,theta2)
%Input 'Conveff' is conversion efficiency or degree of conversion,
%Input 'a' is Vibe constant depends on degree of conversion'Conveff',
%Input 'm' is Vibe form parameter,
%Input 'theta1' is crank angle at start of combustion,
%Input 'theta2' is combustion duration (Start to End of combustion),
%Input 'theta3' is instantaeous crank angle,
%theta1,theta2,theta3 are in degree,
a=-log(1-(Conveff/100));
theta3=0:1:90;
ratio=((theta3-theta1)/theta2);
Tothrr=1-(exp(-a.*(ratio).^(m+1)));
fprintf('\n the vlaue of Tothrr=%f',Tothrr);
plot(theta3,Tothrr,'r-*');
xlabel('Crank angle in Deg','fontsize',12,'fontweight','b');
ylabel('Total Combustion Progression','fontsize',12,'fontweight','b');
grid on;
end

2.3. Program Code- Heat Release Rate


function hrr=calhrr(Conveff,m,mfuel,lhv,theta1,theta2)
%Input 'Conveff' is conversion efficiency or degree of conversion,
%Input 'a' is Vibe constant depends on degree of conversion'Conveff',
%Input 'm' is Vibe form parameter,
%Input 'mf' mass of fuel in gms,
%Input 'lhv' lower heating value of fuel in KJ/Kg,
%Input 'theta1' is crank angle at start of combustion,
%Input 'theta2' is combustion duration (Start to End of combustion),
%Input 'theta3' is instantaeous crank angle,
%theta1,theta2,theta3 are in degree,
a=-log(1-(Conveff/100));
theta3 = 5:1:90;
qftotal=(mfuel/1000)*lhv;
ratio=((theta3-theta1)/theta2);
hrr=qftotal.*a.*(m+1).*(ratio.^m).*exp(-a.*(ratio).^(m+1));
plot(theta3,hrr,'r-*');
xlabel('Crank angle in Deg','fontsize',12,'fontweight','b');
ylabel('Heat Realease Rate in KJ/KW','fontsize',12,'fontweight','b');
legend('Form Factor m = ??')
grid on;
end

With Form factor m = 1.0


With Form factor m = 2.5

With Form factor m = 10.0

CONCLUSION

A single-zone model is often used if there exists a need to have a fast and preliminary analysis of
the engine performance. Single zone models assume that the cylinder charge is uniform in both
composition and temperature, at all time during the cycle. This approach is often used when
simulation is made of a gasoline engine due to the homogeneous combustion. Wiebe function is
one effective approach to model substitute heat release rate is single zone model.
REFERENCES

1. Gunter P.Merker, Christian Schwarz, Gunnar Stiesch, Frank Otto: Simulating


Combustion; Springer-Verlag Berlin Heidelberg 2006, Germany.
2. Fredrik Jones, Christoffer Jezek : Diesel Combustion Modeling and Simulation for
Torque Estimation and Parameter Optimization, Vehicular Systems, 2008.
3. C. E. Goering: ENGINE HEAT RELEASE VIA SPREAD SHEET, 1998 American
Society of Agricultural Engineers 0001-2351 / 98 / 4105-1249, 2008.
4. P.A. Lakshminarayanan, Yogesh V. Aghav: Modelling Diesel Combustion, Springer
Science + Business Media B.V. 2010
1

Combustion Process Characterization in SI engine:

The mass fraction burned profiles as a function of crank angle in SI engines have a characteristic
S-shape. It proves convenient to use these mass fraction burned curves to characterize different
stages of the spark-ignition engine combustion process by their duration in crank angles. The
flame development process, from the spark discharge which initiates the combustion process to
the point where a small but measurable fraction of the charge has burned, is one such stage. The
stage during which the major portion of the charge burns as the flame propagates to the chamber
walls is next. The final stage, where the remainder of the charge bums to completion, cannot as
easily be quantified because energy-release rates are comparable to other energy-transfer
processes that are occurring.
The following definitions are most commonly used to characterize the energy-release aspects of
combustion:

Flame development angle d: The crank angle interval between the spark discharge and the
time when a small but significant fraction of the cylinder mass has burned or fuel chemical
energy has been released. Usually this fraction is 10 percent.
Rapid-burning angle b: The crank angle interval required to burn the bulk of the charge. It is
defined as the interval between the end of the flame-development stage and the end of the flame-
propagation process
Overall burning angle 0: The duration of the overall burning process. It is the sum
of d and b

.
Fig.1 Definition of flame-development angle d and rapid burning angle b, on mass fraction
burned versus crank angle curve.

Figure illustrate these definition on mass fraction burned on fraction of fuel energy released,
versus crank angle plot. While the selection of 10% and 90 % points is arbitrary, such a choice
2

avoids the difficulties involved in determining accurately the shape of curve at the start end of
combustion. These angle can be converted to times (in seconds) by dividing by 6N(with N
revolutions per minute).
A functional form often used to represent the mass fraction burned versus crank angle curve is
Wiebe function:


Xb=1- exp-a m+1

where is the crank angle,0, is the start of combustion, is the total co bustion duration (xb =
0 to xb = I), and a and m are adjustable parametm Varying a and m changes the shape of the
curve significantly. Actual mass fraction burned curves have been fitted with a = 5 and m = 2.

Fig.2 Draw a mass fraction burnt with more rapid combustion process and sketch a
corresponding P- curve.

A useful relationship between the mass fraction burned, xb(mb/m), and the volume fraction
occupied by the burned gas, yb=(Vb/V), can be obtained from the identities
m = mu +mb V= Vu +Vb

and the ideal gas law:

u
Xb=1+ ( - 1)-1
b
3

u
While the density ratio does depend on the equivalence ratio, burned gas fraction in the
b
unburned mixture, gas temperature, and pressure, its value is close to 4 for most spark-ignition
engine operating conditions. Thus, the plot of x, against y, has a universal form, as shown in Fig.

Fig.3 Relation between mass fraction burned xb and volume fracttion burned yb.xr is residual
mass fraction
4

Program for Graph on crank angle Vs % of heat release :


function Tothrr=heatrelease(Conveff,theta1,theta2)
%%% (Conveff is the conversion efficiency)
%%% (theta 1 is the crank angle at the start of combustion)
%%% (theta 2 is total crank angle of combustion)
a=5;
%%% (actual mass fraction burned)
m=2;
theta3=0:1:90;
ratio=((theta3-theta1)/theta2);
Tothrr=1-(exp(-a.*(ratio).^(m+1)));
fprintf('\n the vlaue of Tothrr=%f',Tothrr);
plot(theta3,Tothrr,'r-*');
xlabel('Crank angle in Deg','fontsize',12,'fontweight','b');
ylabel('Total Combustion Progression','fontsize',12,'fontweight','b');
grid on;
clc

end
Output :

Fig4. plot of total combustion progression vs crank angle in deg.


5

Program for graph of burned volume fraction Vs burned mass fraction


yb=0:.01:1;(%%% the volume fraction occupied by burned gas)
z=1+(4*((1./yb)-1))%%%%%%%------For residual Mass Xr=0
xb=1./z;
fprintf('\n the vlaue of xb=%f',xb);
plot(yb,xb,'r-*');
xlabel('burned volume fraction, yb','fontsize',12,'fontweight','b');
ylabel('Burned mass fraction, xb','fontsize',12,'fontweight','b');
grid on;
clc
hold on

z1=1+(3.2*((1./yb)-1))%%%%%%%------For residual Mass Xr=0.2


xb1=1./z1;
plot(yb,xb1,'b-*');

0.9

0.8
xr=0
Burned mass fraction, xb

0.7 xr=0.2

0.6

0.5

0.4

0.3

0.2

0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
burned volume fraction, yb

Fig 5.plot of burned mass fraction vs burned volume fraction.

References:
1.Internal Combustion Engine Fundamentals By John B.Heywood. McGraw-Hill publication
1988 (page no. 389)
IGNITION DELAY IN COMPRESSION IGNITION ENGINES

Arrhenius equation

The Arrhenius type of equation to describe ignition delay is as follows:-



= exp

Where,
= Ignition delay.
= Apparent activation energy for the fuel auto-ignition process.
= Universal gas constant.
and = Constants dependent on the fuel.

Now the representative values for the constants , and are obtained from the experiments in
the constant volume bomb and steady flow reactor which are given in the table 1 below:-

Investigators Conditions Parameters

Apparatus Fuel , atm , K



Spadaccini and Steady flow No.2 Diesel 10-30 650-900 2 2.43e-9 20926
TeVelde No. 1
Spadaccini and Steady flow No.2 Diesel 10-30 650-900 1 4e-10 20080
TeVelde No.2
Stringer Steady flow Diesel 45-50 30-60 770-980 0.757 0.0405 5473
cetane number
Wolfer Constant Fuel with 8-48 590-782 1.19 0.44 4650
volume bomb cetane number
> 50
Hiroyasu Constant Kerosene 1-30 673-973 1.23 0.0276 7280
volume bomb

Ignition delay times calculated with these formulas for various diesel engines are given in the
table 2 below:-

1
Engine Conditions , ms
Speed, P, T, K Spadaccini and Stringer Wolfer Hiroya
rpm atm TeVelde su
No.1 No.2
IDI Diesel
1. Low Swirl 600 45.6 690 17.3 38.2 6.26 3.94 9.60
1200 49.3 747 1.47 3.83 3.22 2.15 3.90
1800 52.5 758 0.86 2.42 2.76 1.82 1.30
2. High Swirl 600 45.2 674 36.3 76.9 7.6 4.68 12.5
1200 48.4 721 4.18 10.3 4.25 2.75 5.67
1800 51.8 744 1.47 4.07 3.19 2.08 3.82
DI Diesel 42.8 781 0.57 1.37 2.60 1.92 2.39
1. Low CR
2. High CR 1500 58.8 975 0.0015 0.0060 0.508 0.407 0.322

From the table we see that, measured ignition delay times in these types of engines are: 0.6 to 3
ms for low compression ratio DI diesel engines over a wide range of operating conditions; 0.4 to
1 ms for high compression ratio and turbocharged DI diesel engines; 0.6 to 1.5 ms for IDI diesel
engines.
The variation in the calculated delay times can be attributed to several factors:

I. In some cases the correlations are being extrapolated outside their original range
of operating conditions.
II. The method used to detect the start of combustion, and hence the duration of
delay, are not identical.
III. The experimental apparatus and the method of fuel-air mixture preparation are
different.

2
Hardenberg and Hase

Figure 3: Comparison of engine ignition delays predicted with equation and measured values

An empirical formula, developed by Hardenberg and Hase for predicting the duration of the
ignition delay period in DI engines, has been shown to give good agreement with experimental
data over a wide range of engine conditions (see Fig. 10-3). This formula gives the ignition delay
(in crank angle degrees) in terms of charge temperature T (kelvins) and pressure p (bars) during
the delay (taken as TC conditions) as:-

. .
(CA) = (0.36 + 0.22 Sp) exp[EA ( )(
.
]

where Sp, is the mean piston speed (meters per second) and R is the universal gas constant
(8.3143 J/mol-K). EA (joules per mole) is the apparent activation energy, and is given by:-

EA =
+

3
where CN is the fuel cetane number. The apparent activation energy decreases with increasing
fuel cetane number. The delay in milliseconds is given by:-

CA
=
0.006

where N, engine speed, is in revolutions per minute. Values for T and p can be estimated using a
polytropic model for the compression process:


= =

where n is the polytropic exponent, rc is the compression ratio, and the subscript i denotes intake
manifold conditions. Values of the polytropic exponent are given in Fig. 10-4 for a direct-
injection diesel under warm and cold engine operating conditions.

4
Figure 4

Physical factors affecting ignition delay


A. Injection timing
i. At normal engine conditions (low to medium speed, fully warmed engine) the
minimum delay occurs with the start of injection at about 10 to 150 BTC.
ii. The increase in delay with earlier or later injection timing occurs because the air
temperature and pressure change significantly close to TC.
iii. With increase in injection advance angle, the reason for increase in delay period is
that the pressures and temperatures are lower when the injection begins.
iv. When the injection advance angles are small, the delay period reduces and
operation of engine is smoother but the power is reduced because larger amount
of fuel burns during expansion (i.e. after burning).
v. Therefore the most favorable conditions for ignition lie in between.

5
B. Injection quantity or Load

Figure 5: Effect of load on ignition delay

Figure 6: Effect of injection quantity on ignition delay

6
i. From the fig-5, we see that the delay decreases approximately linearly with increasing
load for this DI engine.
ii. As the load is increased, the residual gas temperature and wall temperature increase. This
results in higher charge temperature and pressure at injection, thus shortening the ignition
delay.
iii. Fig-6. Shows the influence of injection quantity on the ignition delay. With increase in
injection quantity, the injection pressure increases. This results in smaller drop-sizes that
evaporate quickly and hence in low temperatures. The rate of pre-flame reactions
decreases.Thus, the critical value of the extent of reaction is reached late to increase the
ignition delay.

C. Drop size, Injection velocity, and Rate


These quantities are determined by injection pressure, injector nozzle hole size,
nozzle type and geometry. Experiments by Lyn and Valdmani have shown that none of
these factors has a significant effect on the delay
i. At normal operating conditions, increasing injection pressure produces only
modest decreases in the delay.
ii. In Fig-7, the delay is plotted against injector hole size when injection quantity and
duration were kept constant at 14 mm3and 1.8 ms, with decrease in hole-size
atomization is finer due to high-pressure injection, and hence evaporation of fuel
is intense to cool the air-fuel mixture. The low temperature results in very slow
rate of conversion of fuel to products of combustion i.e. very slow pre-flame
reaction. Therefore, decrease in hole-size increases the time to reach the critical
extent of reaction along the path of drops and hence increase the delay. This effect
is pronounced when the hole-size is between 0.2 and 0.27 mm. As the hole-size is
increased above 0.27 mm, the injection pressure varies very little. Hence, it does
not affect the cooling rate and hence the ignition delay.

7
Figure 7: Effect of injector hole on ignition delay

D. Intake air temperature and pressure

Figure 8: Effect of intake temperature and pressure on ignition delay

8
Fig-8, shows the effects of inlet air pressure and temperature as a function of
engine load.
i. Increase in intake pressure or supercharging reduces the auto-ignition temperature
and hence reduces delay period. Since the compression pressure will increase with
intake pressure, the peak pressure will be higher. Also the power output will be
more as cylinder will contain more air and hence more fuel can be injected per
stroke.
ii. Increasing the intake temperature would result in increase in the compressed air
temperature, which would reduce the delay period. However increasing the intake
temperature, say by preheating, would be most undesirable because it would
reduce the density of air and hence the volumetric efficiency and power output.
This loss would be much greater than the small gain in reduction of delay period.
Preheating the air by 1000C reduces the delay angle by 20.

E. Engine speed
Increase in engine speed at constant load result:-
i. In a slight decrease in ignition delay when measured in milliseconds because the
loss of heat during compression decreases with the result that both the temperature
and pressure of the compressed air tend to rise.
ii. In terms of crank angle degrees, the delay increases almost linearly because the
amount of fuel injected during the delay period depends on crank degrees and not
on absolute time.Since the fuel pump is geared to the engine, at high speeds there
will be more fuel present in the cylinder to take part in the second stage of
uncontrolled combustion resulting in high rates of pressure rise and high maximum
pressures.

Fuel ignition quality


The ignition quality of a diesel fuel is defined by itscetane number. The method used to
determine the ignition quality in terms of cetane number is analogous to that used for

9
determining the antiknock quality of gasoline in terms of octane number. The cetane number
scale is defined by blends of two pure hydrocarbon reference fuels. Cetane (n-hexadecane,
C16H34), a hydrocarbon with high ignition quality, represents the top of the scale with a cetane
number of 100. An isocetane, heptamethylnonane(HMN), which has a very low ignition quality,
represents the bottom of the scale with a cetane number of 15.1 Thus, cetane number (CN) is

CN = percent n-cetane+ 0.15* percent HMN

Ignition quality test diesel fuels

Method 1:
1) The engine used in cetane number determination is a standardized single- cylinder, variable
compression ratio engine with special loading and accessory equipment and
instrumentation.
2) The engine, the operating conditions, and the test procedure are specified by ASTM Method
D613. The operating requirements include: engine speed-900 rev/min; coolant temperature-
1000C; intake air temperature- 65.60C (1500F); injection timing-130BTC; injection pressure-
10.3 MPa.
3) In the test, fuel injection is begun 130 BTC and the compression ratio of the engine is
adjusted till the ignition begins at exactly TDC, i.e. an ignition delay of 130 is achieved.
4) The actual numerical values are set by means of reference fuel.
5) The cetane number of a fuel is the percentage by volume of cetane in a mixture of cetane and
heptamethylnonane that has the same performance in the standard test engine as that of the
fuel.

Method 2:

1) This method is based on the diesel index which is defined as:-


Diesel index = (Aniline point (0F) API Gravity at 600F (150C))/100
2) It is applicable to straight run fuels, catalytically cracked stocks and blends of the two. Its
use is suitable for most diesel fuels and gives numbers that correspond quite closely to
cetane number.

10
3) Aniline point is the lowest temperature at which oil is completely miscible with an equal
volume of aniline. For a good quality diesel oil the aniline point is greater than 700F.
4) API (American Petroleum Institute) gravity is given by:-

API = (141.5/specific gravity at 600F (150C)) 131.5

Effect of fuel properties

Figure 9: Effect of fuel molecular structure

The ignition quality of the fuel, defined by itscetane number, will obviously affect the delay. The
dependence ofcetane number on fuel molecular structure is as follows:-

i. Straight-chain paraffinic compounds (normal alkanes) have the highest ignition quality,
which improves as the chain length increases.

11
ii. Aromatic compounds have poor ignition quality as do the alcohols (hence, the difficulties
associated with using methanol and ethanol, possible alternative fuels, in compression-
ignition engines).
iii. Figure-9 illustrates these effects. A base fuel was blended with pure paraffinic (normal,
iso-, and cycloalkanes), aromatic, and olefinic hydrocarbons, of various carbon numbers,
by up to 20 percent by volume.
iv. The base fuel, a blend of 25 percent n-hexadecane and 75 percent isooctane, had a cetane
number of 38.3. The figure shows that the resulting ignition delays correlate well as a
function of cetane number at constant compression ratio and engine operating conditions.
v. Addition of normal alkanes (excluding n-pentane and lower carbon number alkanes)
improves the ignition quality. As the chain length of added paraffin gets longer (higher
carbon number) the cetanenumber improvement increases.
vi. Isoalkanes, depending on the degree of branching, degrade ignition quality (unless the
branching is concentrated at one end of the molecule, when these types of isoalkanes
improve ignition quality).
vii. Cycloalkanes and aromatics generally reduce the cetane number, unless they have a
long n-alkane chain attached to the ring.
viii. The cetane number of a fuel (a measure of its ability to autoignite) generally varies
inversely with its octane number (a measure of its ability to resist autoignition; on
knock). The cetane number of commercial diesel fuel is normally in the range 40 to 55.

12
IGNITION DELAY IN CI ENGINE

The period between the start of fuel injection into the combustion chamber and the start of
combustion is termed as ignition delay period.

Combustion phases in C.I. Engine:


1. Ignition delay period
2. Premixed combustion phase
3. Mixing-controlled combustion phase
4. Late combustion phase

The start of combustion is determined from the change in slope on the p- diagram as may be
seen from the figure below, or from heat release analysis of the p() data, or from luminosity
detector in experimental conditions. Start of injection can be determined by a needle-lift
indicator to record the time when injector needle lifts off its seat.
Start of combustion is more difficult to determine precisely. It is best identified from the
change in slope of heat release rate, determined from cylinder pressure data.

In DI engines, ignition is well defined, in IDI engines ignition point is harder to identify.
Both physical and chemical processes must take place before a significant fraction of the
chemical energy of the injected liquid fuel is released.

Ignition delay is comprised of two processes and how these processes affect the engine
design and operating variables, fuel characteristics are as follows:

1. Physical Process
i. Atomization of liquid fuel jet: Good atomization requires high fuel injection
pressure, small injector hole diameter, optimum fuel viscosity and high
cylinder air pressure at the time of injection.
ii. Rate of vaporization of fuel droplets: it depends on the size of the droplet,
their distribution, and their velocity, the pressure and temperature inside the
chamber, and volatility of the fuel.
iii. The rate of fuel air mixing is controlled largely by injector and combustion
chamber designs. Engine designs use a pre-chamber or a swirl chamber to
create the vigorous air motion necessary for rapid fuel air mixing while
injector design features such as the number and spatial arrangement of the
injector holes.

2. Chemical Process:
The chemical processes are the pre-combustion reactions of the fuel, air, residual
gas mixture which lead to auto-ignition.

Chemical delay is more effective for the duration of the ignition delay period.

Cetane Number or CN is a measure of a fuel's ignition delay, the time period between the
start of injection and the first identifiable pressure increases during combustion of the fuel. In
a particular diesel engine, higher cetane fuels will have shorter ignition delay periods than
lower cetane fuels.

Generally, diesel engines operate well with a CN from 40 to 55. Fuels with higher cetane
number have shorter ignition delays, providing more time for the fuel combustion process to
be completed. Hence, higher speed diesel engines operate more effectively with higher cetane
number fuels.

Physical Factors Affecting Ignition Delay:

Injection Timing-
At normal operating conditions min ignition delay (ID) occurs with start of injection at 10 to
15 BTDC. Cylinder temperature and pressure drops if injection is earlier or later (high at
first but decreases as delay proceeds)

Injection quantity (load)-


Reducing engine load changes AFR, cools down the engine, reduces wall temperatures,
reduces residual gas temperatures and increases ID

Droplet size, injection velocity and rate-


Ignition quality within practical limits do not have significant effect on ID
Increases in injection p produces only modest decreases in ID
Injector nozzle diameter effects droplet size but has no significant effect on ID

Combustion chamber design-


Spray impingement on the walls effect fuel evaporation and ID
Increases in compression ratio, increases p and T and reduces ID
Reducing stroke volume, increases surface area to volume ratio, increases engine cooling and
increases ID
Swirl Rate:
Change evaporation rate and air-fuel mixing - under normal operating conditions the effect is
small.
At start-up (low engine speed and temperature) more important, high rate of evaporation and
mixing is obtained by swirl
Oxygen Concentration:
Residual gases reduce O2 concentration and reducing oxygen concentration increases ID
Cetane Number: Both physical and chemical properties of the fuel is important. Ignition
quality of the fuel is defined by its cetane number.

Straight chain parafinic compounds (normal alkanes) have highest ignition quality, which
improves as the chain length increases.

Aromatic compounds, alcohols have poor ignition quality. Cetane number can be increased
by ignition-accelerating additives like organic peroxides, nitrates, nitrites and various sulphur
compounds. Most important (commercially) is alkyl nitrates about 0.5% by vol in a
distillate fuel increases CN by 10.

Normal diesel fuel has CN of 40 to 55 (high speed 50 60, low speed 25 45)

ANALYSIS-

Hardenberg and Hase expressed ignition delay for naturally aspirated or turbocharged
diesel engine on the basis of crank angle (CA) and milliseconds (ms) as
Where,
i is ignition delay,
N is rpm,
MPS is mean piston speed in m/s,
R is the universal gas constant (8.31434 J/mole),
Tm is the mean temperature in the inlet manifold in K,
rc the compression ratio, and n is the polytropic index of compression.
The activation energy EA is given by

Where CN = Cetane number

1. Mean piston speed = f(Stroke length, Speed)


2. Activation Energy Ea = f(Cetane Number)
3. Temperature = f(Polytropic index, compression ratio)
4. Pressure = f(Polytropic index, compression ratio)

Output Variables Input Variables Input Parameter

ID in CA (deg) N (rpm) Compression Ratio


ID in ms Inlet Temp. (K) Stroke Length
Inlet Pressure (in Pa) Cetane Number
Temperature Vs ID in CA (with varying compression ratio)

Inlet temperature Vs ID in CA (with varying pressure)

Variation of Ignition delay with CR


Conclusion

1. For higher Compression Ratios, which imparts favourable combustion conditions like
better automization, higher temperatures at the end of compression leads to reduced
delay period.

2. For higher inlet temperature conditions ignition delay is shorter.

3. Cylinder pressure has, but a small effect on ID.


Mathematical Models Of Diesel Engine Combustion using Wiebe
Function Analysis

1
= ( + )
1 1
Where the terms on the left and right are as below;


Represents the heat release rate


Represents heat loses. Heat transfer loses can be calculated by using

Woschis or Annands correlation.

Assumptions:

1. Sensible enthalpy of the injected fuel and the variation of Cv with T has been
neglected
2. The specific heat ratio can be calculated from thermo chemical data, assuming
that the mixture composition is fixed by simple stoichiometry and linearly related
to the degree of reaction

Inputs Needed:

1. Above equation can be used to perform heat release analyses if the pressure
diagram is known.

It can also be used to predict the cylinder pressure and temperature if the heat release rate
dQ/d is secified.

In order to describe the premixed and diffusive combustion period observe in diesel
engines, two Wiebe function can be used as follows.
+1

= 6.9 ( + 1) exp [6.9 ]


+1
+ 6.9 +1 ( ) exp [6.9 ( ) ]

Where;

1. The subscripts p and d refer to premixed and diffusive combustion respectively;


2. Mp and Md are shape factors corresponding to premixed and diffusive combustion
respectively;
3. p and d are the duration of the energy release in premixed and diffusive
combustion respectively;
4. Qp and Qd characterizes the premixed and diffusive combustion respectively.
Typical values of premixed and diffusive combustion parameters are shown in
figure.

Fig. 3.15 shows the combustion rate computed with two wiebe function Eq 59 as a
function of crank shaft angle for three different angles and fig.3.16 shows schematic
premixed of combustion parameters Qp and the amount of fuel injected during ignition
delay time QI .

Fig. 3.16 also shows that Qp =0.5QI for a variety of fuels in DI diesel engine. Fig 3.16
indicates that the diffusive combustion parameters are functions of the injection timing
Miyamoto performed experiments in DI and IDI diesel engines for a range of
speeds , loads injection timings, fuels and steady transient conditions; they anslysed the
conbustjon data by measns of double Wiebe function Eq 59 .

Figure :- Experimental (dashed lines) and predicted (solid lines) heat release rates
in a diesel engine for three different injection timing.

They found that Mp, Md and d remained constant in both DI and IDI engines (Mp
= 3 for DI and IDI engines ; Md = 0.5 for DI engines; Md = 0.9 for IDI engines; and d
=7 CA (crank angle degree) for DI and IDI engines).
Figure :- Correlation between the premixed Qp combustion parameters and the enthalpy
of the fuel injected during the injection delay time Ql (left) and the schematic of Qp and
Ql.(left)

However, the values of Qp, Qd and d were found to depend on the fuel injection
rate and ignition delay time. The premixed combustion parameter Qp correlated linearly
with the amount of fuel injected during the ignition delay time, and the duration of the
diffusive burning d was a linear function of the diffusive combustion parameter.

Qd.

The engine, smoke, noise, and thermal efficiency were found to correlate with Qp
and p. Both the smoke and the noise can be reduced by decreasing Qp while maintaining
a value of d less than 60 CA.

Ghojel used Eq. (58) and to wiebe functions similar to Eq. (59) to predict the
pressure diagram in DI diesel engine.

Shipinski, Sitkei, and Tinaut used a single exponential Wiebe function to predict
the average pressure and temperature in DI and IDI engines. In these studies the burning
rate was expressed as;
+1
,
= +1 ( ) exp [ ( ) ]
b

b = end ig

Figure :- correlation between the diffusive combustion duration parameter d and the
diffusive combustion parameter Qd in direct injection and direct injection diesel engines.
Figure:- Dependence of the premixed and diffusive combustion parameter on the
injection timing in DI and IDI diesel engines.

Where Hc is the enthalpy of combustion, mf,inj is the total amount of fuel injected
into the cylinder each cycle, ig is the crank shaft at the start of combustion, is the crank
shaft angle, and a and m are the wiebe function parameters. The value of parameter a
depends on the fuel air ratio; b depand on the equivalence ratio and rpm; and m
depends on engine rpm and the pressure and temperature at intake valve closing.

The model presented in section 3.2.3 have been used analyze the
transient performance of turbocharged diesel engine [60], IDI diesel engines(sigle-zone
model with experimentally determined heat release rates for IDI diesel engines)[61],two
stroke turbocharged diesel engine[63],and turbocharged turbocompouned diesel
engine[19].

Watson[60] used the empirical fuel burning rate Watson et al.and sub models for the
turbo charger, charge air cooler, friction, governor and fuel controller to analyze the
transient performance of turbocharged diesel engines. However, the parameter that
appear in the Wiebe functions [Eq, 59 and 60] are the function of the engine operating
conditions(rpm.load),combustion chamber geometery etc,

Brandstatter used two Wiebe functions to analyze combustion in a prechamber


stratified-charge engine with injection in the prechamber. He included the flows through
the valves,intake and exhaust systems, used quasi steady adiabatic flow expressions to
calculate the mass flow rate through the throat that separates the prechamber and main
chamber, simulated the entire engine cycle, assumed perfect mixing of the burnt and
unburnt gases in the prechamber and main chamber but did not consider wave
phenomena in the exhaust and intake manifolds. He also used real gas effects and
considered that the mixture sp.heat at constant pressure is a function of the pressure,
temperature, and equivalence ratio.

Matlab Program For Heat Release Rate Estimation

clear all
clc

Mp = 3;
Md = 0.9;
Qp = 100;
Qd = 400;
ao = 0;
ap = 7;
ad = 60 ;

A1 = linspace(ao,ap,180);
A2 = linspace(ap, ad, 180);
A = pi/180*A1;
An = pi/180*A2;
Ap = pi/180*ap;
Ad = pi/180*ad;

z1 = ((6.9.*Qp./Ap).*(Mp + 1).*(A./Ap).^Mp.*exp(-6.9.*(A./Ap).^(Mp+1)))
+ ((6.9.*Qd./Ad).*(Md + 1).*(A./Ad).^Md.*exp(-6.9.*(A./Ad).^(Md+1)));

z2 = ((6.9.*Qd./Ad).*(Md + 1).*(An./Ad).^Md.*exp(-
6.9.*(An./Ad).^(Md+1))) + ((6.9.*Qp./Ap).*(Mp + 1).*(An./Ap).^Mp.*exp(-
6.9.*(An./Ap).^(Mp+1)));

plot(A1, z1, '-b'); grid on;


hold on;
plot(A2, z2, '-r');
xlabel('Crank Angle (Deg)');
ylabel('Heat Release Rate (dq/d\theta)');
Mathematical Models of Diesel Engine Combustion using Wiebe
Function Analysis

Whitehouse Way Model


In this model atomization of fuel droplets, evaporation of fuel, entrainment of air,
and micromixing of fuel and air are collectively known as preparation of fuel known.

If the burning time is negligible compared with the rate of fuel preparation, the
hat release rate can be considered as a measure of the fuel preparation rate for most of the
burning period. Chemical kinematics (Kinetics) is a rate-controlling process of the
beginning of burning when the temperature is low. As, the temperature increases, the
burning rate and heat release rate increase, but they decreases as the prepare fuel is
deleted.

In the model prepared Whitehouse and Way combustion is assumed to occur at


the droplet surface. If the droplets are of uniform diameter, the rate of fuel preparation is
given by the following expressions


=

Where;

minj is the mass of the fuel injected into the cylinder;

mup is the mass of injected unprepared fuel;

po is the oxygen partial pressure;

a1 = 1/3;

a2 =2/3;

m = 0.4 and

The values of c and m need to be adjusted in order to achieve good agreement with
experimental data for a given engine.

Eq. 62 indicates that the fuel preparation rate depends on the availability of the oxygen in
the cylinder.

During premixed combustion the burning rate is controlled by the chemical kinetics, and
during the period the fuel burning rate is given by the following equation


= ep ( )
0.5

fbr = (c/N) * (po/(T^0.5)) * exp(-E/(RT)) (Above Difference)

Where;

c' and E/R are constants that control ignition delay period and

mfb denotes the burnt fuel.

The fuel burning rate is less than the preparation rate at the start of combustion.
This results in accumulation of the prepared fuel, which causes an increase in the burning
rate.

Figure: Rates the fuel preparation and heat release

Fig. shows the fuel burning rate as a function of the crank shaft angle. The first peak in
the figure corresponds to the combustion of premixed fuel and air. The decrease in the
burning rate is due to the depletion of prepared fuel. After the prepared fuel is exhausted,
the heat release rate is equal to the fuel preparation rate.

Although no ignition delay time is considered in the model prepared by


Whitehouse and Way, the small burning rate during the initial stages of combustion can
be considered to represent the delay time. Thus, c and E/R effectively control the delay
period.

The Whitehouse-Way model is consistent with experimental observations


showing that the major factors affecting the burning rate are the injection rate, engine
speed, and combustion chamber design, all of which affect the fuel air mixing.

The fuel air mixing affects the burning process in a diesel engine. These processes
proceed in the three phases. In the first phases the premixed part of the fuel burns, and the
burning rate is highest in typical DI engine. In the second phase the heat is released in a
crank angle interval, which may be eight times greater than that of the premixed burning.
The burning rate decays, and the fuel jet burns as a highly turbulent diffusion flame in the
second phase. The burning rate is quite low in the third phase, which may last through the
expansion stroke. The phase account for about 10% to 20% of the total heat input. The
nature of the combustion in the third phase is not clear, however but appears to be an
inherent part of diesel engine combustion.

Recently, Carapanayotis and Salcuden used a modifies version of the


Whitehouse-Way model that accounts for the ignition delay time, fuel vaporization, and
combustion as follows. The ignition delays ways experimentally determined as a function
of the engine rpm and cylinder pressure at the beginning of the injection and the
following expression was obtained:

td = 2.5 p-0.009x (s) x = (rpm)1/2

The amount of fuel vapor in the cylinder at any crank angle after the start of injection was
calculated, assuming that the spray was characterized by a single droplet of sauter mean
diameter equal to SMD. For a single droplet in the absence of combustion, the following
expression apply:


d2 = SMD2 -
ln (1 + B)


B = Cp
) +1
Where L is the latent heat of vaporization, T is the gas temperature far from the droplet,
Ts is the temperature at the droplet surface, l is the liquid droplet density, s is the
density at the doplet surface, and D is the mass diffusivity.

During the ignition delay period, some of the injected fuel is evaporated and
mixed with air, forming the combustible mixture. In the early stages, combustion occurs
under premixed conditions at a rate equal to

RRp = 105 Y1.2o Y1.1 f exp (-2000/T) (kg/CA)

Where Yo and Yf = The mass fraction of oxygen and fuel vapor, respectively,

RRp= The premixed reaction rate

Equation (67) is assumed to be valid until the amount of fuel evaporated at the
end of the ignition delay period has been consumed, RRp = 0 and combustion is assumed
to be controlled by the fuel air mixture rate, according to the following expression

RRm = 0.017 minj Po

minj = mass of fuel injected at the crank angle

Po = Oxygen partial pressure

RRm = Mixing controlled reaction rate

The mass burning rate during the premixed and diffusive periods can be expressed
as

BR = RRp + RRm

BR = Burning rate, during premixed combustion RRp >> RRm

Carapanayotis and Salcuden [68] used Woschis heat transfer correlation with

mot = 6.18 pis comp = 0 during scavenging

mot = 2.28 pis comp = 0 during compression


1

COMBUSTION STOICHIOMETRY

Introduction:

Combustion is the conversion of a substance called a fuel into chemical


compounds known as products of combustion by combination with an oxidizer. The
combustion process is an exothermic chemical reaction, i.e., a reaction that releases
energy as it occurs. Thus combustion may be represented symbolically by:

Fuel + Oxidizer Products of combustion + Energy

Here the fuel and the oxidizer are reactants, i.e., the substances present before the
reaction takes place. This relation indicates that the reactants produce combustion
products and energy.In combustion processes the oxidizer is usually air but could be pure
oxygen.Combustion of the fuel-air mixture inside the engine cylinder is one of the
processes that controls engine power, efficiency, and emissions.

Combustion Stoichiometry:

This section develops relations between the composition of the reactants (fuel and
air) of a combustible mixture and the composition of the products. Since these relations
depend only on the conservation of mass of each chemical element in the reactants, only
the relative elemental composition of the fuel and the relative proportions of fuel and air
are needed.

If sufficient oxygen is available, a hydrocarbon fuel can be completely oxidized.


The carbon in the fuel is then converted to carbon dioxide CO2 and the hydrogen to water
H20. For example, consider the overall chemical equation for the complete combustion of
one mole of propane C3H8:

C3H8 + aO2 bCO2 + cH2O (1)


2

A carbon balance between the reactants and products gives b = 3. A hydrogen


balance gives 2c = 8, or c = 4. An oxygen balance gives 2b + c = 10 = 2a, or a = 5. Thus
Eq. (1) becomes

C3H8 + 5O23CO2 + 4H2O(2)

Note that Eq. (2) only relates the elemental composition of the reactant and
product species; it does not indicate the process by which combustion proceeds, which is
much more complex. Air contains nitrogen, but when the products are at low
temperatures the nitrogen is not significantly affected by the reaction.

Let us represent the chemical formula of a fuel as CaHbOcNd. A stoichiometric


reaction is defined such that the fuel burns completely and the only products are carbon
dioxide and Water. Let us write the following stoichiometric reaction

CaHbOcNd + A(O2 + 3.76N2) n1 CO2 + n2 H2O + n3 N2 (3)

and solve for A, the stoichiometric molar air-fuel ratio, and the moles ni, (i =
1,2.3) that describe the product composition. Note that the stoichiometric reaction is
expressed per mole of fuel.

Note that only the ratios of the numbers in front of the symbol for each chemical
species are defined by Eq. (3); i.e., only the relative proportions on a molar basis are
obtained. Thus the fuel composition could have been written CaHb, where y = b/a.

Figure 1: Stoichiometric A/F ratio for air-hydrocarbon fuel mixtures as a function of H/C ratio
3

We know that atoms arc conserved, so we can write

C: =

H: =

O: + = + (4)

N: + . =

Solution of these four equations gives

=( + )

= =

= ( ) + . ( + )

The stoichiometric air-fuel ratio S is

[ . . ]
= (6)
. + . + . + .

The molecular weights of oxygen, atmospheric nitrogen, atomic carbon, and


atomic hydrogen are, respectively, 32, 28.16, 12.011, and 1.008. (A/F)s depends only on
y.

Example 1. A hydrocarbon fuel of composition 84.1 percent by mass C and 15.9 percent
by mass H has a molecular weight of 114.15. Determine the number of moles of air
required for stoichiometric combustion and the number of moles of products produced
per mole of fuel. Calculate (A/F)s, (F/A)s, and the molecular weights of the reactants and
the products.

Solution:

Assume a fuel composition CaHb . The molecular weight relation gives

. = . + .
4

The gravimetric analysis of the fuel gives

/ = . / . / . / .

= .

Thus

= & =

The fuel is octane C8H18, . Equation (3) then becomes

+ . + . + + .

In moles:

+ . + . = + + .

+ . = .

Relative mass:

. + . . = . + . + . .

. + . = .

per unit mass fuel:

+ . = .

Thus for stoichiometric combustion, 1 mole of fuel requires 59.66 moles of air
and produces 64.16 moles of products. The stoichiometric (A/F)s is 15.14 and (F/A)s is
0.0661. The molecular weights of the reactants MR and products MP are

= = . + . .
.

= = . + . + . .
.

= .

= .
5

Fuel-air mixtures with more than or less than the stoichiometric air requirement
can be burned. With excess air or fuel-lean combustion, the extra air appears in the
products in unchanged form. For example, the combustion of iso-octane with 25 percent
excess air, or 1.25 times the stoichiometric air requirement, gives

+ . . + . + + . + . (7)

With less than the stoichiometric air requirement, i.e., with fuel-rich combustion,
there is insufficient oxygen to oxidize fully the fuel C and H to CO2 and H2O. The
products are a mixture of H2O with carbon monoxide CO2 and CO and hydrogen H2 (as
well as N2). The product composition cannot be determined from an element balance
alone and an additional assumption about the chemical composition of the product
species must be made. (Because the composition of the combustion products is
significantly different for fuel-lean and fuel-rich mixtures, and because the stoichiometric
fuel/air ratio depends on fuel composition, the ratio of the actual fuel/air ratio to the
stoichiometric ratio (or its inverse) is a more informative parameter for defining mixture
composition: The fuel/air equivalence ratio ,

= / / / (8)

will be used throughout this text for this purpose. The inverse of, the relative
air/fuel ratio ,

= = / / / (9)

is also sometimes used.

For fuel-lean mixtures: <1, >

For stoichiometric mixtures: = = 1

For fuel-rich mixtures: >1, <1

When the fuel contains oxygen (e.g., with alcohols), the procedure for deter-
mining the overall combustion equation is the same except that fuel oxygen is included in
6

the oxygen balance between reactants and products. For methyl alcohol (methanol),
CH3OH, the stoichiometric combustion equation is

+ . + . + + . (10)

and(A/F)s = 6.47. For ethyl alcohol (ethanol), C2H5OH, the stoichiometric


combustion equation is

+ + . + + . (11)

And (A/F)s =9.00.

If there are significant amounts of sulfur in the fuel, the appropriate oxidation
product for determining the stoichiometric air and fuel proportions is sulfur dioxide, SO2.

For hydrogen fuel, the stoichiometric equation is

+ + . + . (12)

and the stoichiometric (A/F) ratio is 34.3.

Note that the composition of the products of combustion in Eqs. (7) and (10) to
(12) may not occur in practice. At normal combustion temperatures significant
dissociation of CO2 and of H2O occurs. Whether, at low temperatures, recombination
brings the product composition to that indicated by these overall chemical equations
depends on the rate of cooling of the product gases.

At low temperatures (T < 1000 K, such as in the exhaust) and carbon to oxygen
ratios less than one, the overall combustion reaction for any equivalence ratio can be
written

+ / + .

+ + + + + (13)
7

For reactant C/O ratios greater than one we would have to add solid carbon C(s) several
other species to the product list, as we shall see. This equation assumes the dissociation
of molecules is negligible.

For lean combustion ( < 1) products at low temperature, we will assume no


product CO and H2, i.e., n5= n6 = 0. In this case atom balance equations are sufficient to
determine the product composition since there are four equations and four unknowns.

For rich combustion we will assume that there is no product O2, i.e., n4 = 0. In
this case there are five unknowns, so we need an additional equation to supplement the
four atom balance equations. We assume equilibrium considerations between the product
species CO2, H2O, CO, and H2 determine the product composition. This reaction is
termed the water-gas reaction:

+ + (14)

with the equilibrium constant providing the fifth equation:


= (15)

The equilibrium constant K(T) equation is a curve fit of JANAF Table data for 400< T <
3200 K:

. . .
= .

+
(16)

= /

In the rich case, the parameter n5 is given by the solution of the quadratic equation

+
= ( )(17)

Species ni 1 > 1
8

CO2 n1 A a n5

H2O n2 b/2 b/2 d1 + n5

N2 n3 (d/2) + 3.76 A/ (d/2) + 3.76 A/

O2 n4 A(1/ d1 - 1) 0

CO n5 0 n5

H2 n6 0 d1 n5

Table 1: Low Temperature Combustion Products

The a1, b1, c1, d1 coefficients are given by

= + . (18)

= . .

In reciprocating engines there is residual gas mixed with the fuel and air. We need
to determine the composition of such a fuel-air-residual gas mixture for analysis of the
compression stroke and later the unburned mixture ahead of the flame. The residual gas is
assumed to be at a low enough temperature so that the species relations in Table 1 specify
its composition.
9

Figure 2: Equilibrium composition of octane (C8H18)-air mixture at T=3000K, P=50bar

Example 2. What is the mole fraction (ppm) of CO produced when C8H18 is burned at
= 1.2 and T = 1000 K?

Solution:An equivalence ratio of = 1.2 is a rich mixture. so the productconcentration of


O2 is assumed to be zero. i.e., n4= 0. The calculation proceeds as fol1ows:

= ,

= ,
10

= =

=( + )= .

= ( )= .


= =

. . .
= ( . + )= .

= = .

= + . = .

= . . = .

+
= = .

= = .

= + = .


= ( ) + . ( )= .

= = .

= = .

The reaction equation is


11

+ . + .
. + . + . + . + .

and

mole fraction of CO = 0.03009 30,000 ppm.

References:

1) Internal Combustion Engines, by J. B. Heywood.


2) Internal Combustion Engines, by C. R. Ferguson.
12

MATLAB PROGRAMS:

1) To determine Stoichiometric A/F ratio of given fuel

function As = stoichiometric(a, b, c, d)

AIR = (a+(b/4)-(c/2)); %moles_of_AIR


moles_of_CO2 = a;
moles_of_H2O = (b/2);
moles_of_N2 = (d/2)+3.76*(a+(b/4)-(c/2));

As=(28.85*(4.76*AIR)/(12.01*a + 1.008*b + 16*c +14.01*d)); %calculation


of stoichiometric ratio of fuel
fprintf ('\nStoichiometric A/F ratio of fuel=%f\n',As)
fprintf ('\nMoles of AIR required for combustion of 1mole of fuel=%f\n
Moles of CO2 evolved=%f \nMoles of H2O evolved=%f\nMoles of N2
evolved=%f\n ',AIR,moles_of_CO2,moles_of_H2O,moles_of_N2)
end

2) To determine equilibrium constant

function K = equilibriumK(T)

t = T/1000;
K = (2.718^((2.743)-(1.761/t)-(1.611/t^2)+(0.2803/t^3)))

end

3) To plot the graph of (A/F)S Vs H/C

y=0:0.1:2;%y(i)=b/a;
fori=1:1:length(y)
Si(i)=(34.56*(4+y(i)))/(12.011+1.008*y(i));
end
holdon
13

figure(1)
plot(y,Si,'-r')
title('Stoichiometric A/F ratio for air-hy(i)drocarbon fuel mixtures as
a function of molar H/C ratio');
xlabel('Fuel molar H/C ratio');
ylabel('(A/F)s');

Figure 3:Stoichiometric A/F ratio for air-hydrocarbon fuel mixtures as a function of H/C ratio

4) To determine mole fractions for 1

function L = leaner(a, b, c, d, phi)


phi=0:0.01:1;
fori= 1:1:length(phi)
if phi<=1 % if mixture is leaner or stoichiometric
AIR=(a+(b/4)-(c/2));
14

CO2=a;
H2O= (b/2);
N2=(d/2)+3.773.*((AIR)./phi);
O2=AIR.*((1./phi)-1);
CO=0;
H2=0;

TM= (CO2+N2+O2+H2O); %total moles of products


MF_CO2 = CO2./TM; %mole fraction of CO2
MF_H2O = H2O./TM; %mole fraction of H2O
MF_N2 = N2./TM; %mole fraction of N2
MF_O2 = O2./TM; %mole fraction of O2

figure(1)
semilogy(phi, MF_CO2)
text(phi(end), MF_CO2(end),'CO2')
holdon
semilogy(phi,MF_H2O,'r')
text(phi(end),MF_H2O(end),'H20')
semilogy(phi,MF_N2,'g')
text(phi(end),MF_N2(end),'N2')
semilogy(phi,MF_O2,'k')
text(phi(end),MF_O2(end),'O2')

axis([0 1 .0001 1])


title('mole fraction vs phi for (phi<=1)')
xlabel('phi');
ylabel('mole fractions');
end
end
end
15

Figure 5: Mole fraction Vsfor 1

5) To determine mole fraction for > 1

function R = richer(a, b, c, d, T)
K = equilibriumK(T);
phi=1:0.01:10;
for i= 1:1:length(phi)
AIR=(a+(b/4)-(c/2));%moles of AIR

d1=2.*AIR.*(1-(1./phi));
c1=-a*d1*K;
b1=(b/2)+(K*a)-(d1*(1-K));
a1=1-K;
16

n5=(-b1+(((b1.^2)-(4*a1*c1)).^0.5))./(2*a1);

% MOLES

CO=n5; %moles of CO
CO2=a-n5; %moles of CO2
H2O=(b/2)-(d1)+(n5); %moles of H2O
N2=d./2+(3.76.*AIR)./phi; %moles of N2
O2=0; %moles of O2
H2=d1-n5; %moles of H2

% MOLE FRACTIONS

TM= (CO2+N2+H2O+CO+H2); %total moles of products


MF_CO = CO./TM; %mole fraction of CO
%fprintf('\n mole fraction of CO=%f',MF_CO)
MF_CO2 = CO2./TM; %mole fraction of CO2
%fprintf('\n mole fraction of CO2=%f',MF_CO2)
MF_H2O = H2O./TM; %mole fraction of H2O
%fprintf('\n mole fraction of H2O=%f',MF_H2O)
MF_N2 = N2./TM; %mole fraction of N2
%fprintf('\n mole fraction of N2=%f',MF_N2)
MF_H2 = H2./TM; %mole fraction of H2
% fprintf('\n mole fraction of H2=%f',MF_H2)

% graph

figure(3)

loglog(phi, MF_CO2,'g')
text(phi(end), MF_CO2(end),'CO2')
holdon
loglog(phi,MF_H2O,'r')
text(phi(end),MF_H2O(end),'H20')
loglog(phi,MF_N2,'g')
text(phi(end),MF_N2(end),'N2')
loglog(phi,MF_H2,'b')
17

text(phi(end),MF_H2(end),'H2')
loglog(phi,MF_CO,'k')
text(phi(end),MF_CO(end),'CO')

axis([1 10 .001 1])


title('mole fraction vs phi for (phi>1)')
xlabel('phi');
ylabel('mole fractions');
end
end

Figure 6: Mole fraction Vsfor> 1


Cetane number and ignition delay prediction of diesel and diesel blend fuels

1. Introduction

Cetane number (CN) is a measure of the ignition quality of the diesel fuel and is
determined by a standard engine test as specified by ASTM (ASTM D613). The ignition
quality is quantified by measuring the ignition delay, which is the period between the time of
injection and the start of combustion (ignition) of the fuel. A fuel with a high CN has a short
ignition delay period and starts to combust shortly after it is injected into an engine. The
ignition quality of the diesel fuel depends on its molecular composition. Some of the simpler
molecular components such as the n-paraffins can ignite in a diesel engine with relative ease,
but others like aromatics have more stable ring structures that require higher temperature and
pressure to ignite.
The ASTM D6131 test method defines the CN of a diesel fuel as the percentage by
volume of normal cetane (C16H34), in a blend with 2,2,4,4,6,8,8-heptamethylnonane
(sometimes called HMN or isocetane), which matches the ignition quality of the diesel fuel
being rated under the specified test conditions. By definition, normal cetane has been
assigned a CN of 100, whereas HMN has a CN of 15. The ASTM D613 method involves
running the fuel in a single-cylinder compression ignition engine with a continuously variable
compression ratio under a fixed set of conditions. Although it has been the traditional test
method for CN, it was soon recognized that the test suffered from many disadvantages, some
of which include a relatively large fuel sample volume requirement (1 L), significant time
consumption (approximately a few hours), and a relatively high reproducibility error. ASTM
documents a maximum reproducibility error of 3-4 numbers for this test. Consequently, there
have been many attempts to develop alternate tests to replace the ASTM D613 method. These
include devising better engine tests and developing correlative models to predict the CN from
bulk properties of the fuel that may be measured more quickly and reliably. Among the
different engine tests developed, the Ignition Quality Tester (IQT) has been the most
promising for reliable measurement of CNs of middle distillates and alternate fuels. It uses
much smaller sample volumes (<50 mL) and requires much shorter times (<10 min) for
completion of the test. The IQT consists of a constant volume combustion chamber that is
electrically heated, a fuel injection system, and a computer that is used to control the
experiment, record the data, and provide interpretation of the data. The basic concept of the
IQT is the measurement of the time from the start of fuel injection into the combustion
1
chamber to the start of combustion. The thermodynamic conditions in the combustion
chamber are influenced by the charge pressure (combustion air), chamber temperature, and
fuel injection quantity. These conditions are precisely controlled to provide consistent
measurement of the ignition delay time. The measured ignition delay time is correlated by an
equation to a derived CN. However, in terms of actual numbers, the IQT-based derived CN is
the same as what we would have measured if we had run the fuel through the actual ASTM
engine. The reproducibility errors on IQT based derived CN are tighter than the those of
ASTM D613 and vary between 1 and 2 numbers. The present work is therefore based on IQT
measurements on the various diesel fuels and their commercial blends.
2. Methods for predicting cetane number
Numerous attempts have been made in the past to correlate the CN with various
physical and chemical attributes of the diesel fuel. These include correlations based on either
bulk properties such as API gravity, boiling points, and aniline points or based on molecular
composition, measured by gas chromatography (GC) or inferred from different spectroscopic
methods such as Fourier transform infrared spectroscopy, nuclear magnetic resonance (1H
NMR, 13C NMR), etc. The D4737 equation correlates the API gravity and the T10, T50, and
T90 points on the boiling curve to the CN of the fuel and is summarized as follows[1]:
CN 45.2 0.0892T 10N (0.131 0.901BN )T 50N (0.0523 .42 BN )T 90N
[0.00049(T 10N ^ 2 T 90N ^ 2)] 107 BN 60( BN ^ 2)
where,
d = specific gravity at 60 F,
BN = e-3.5(d-0.85) - 1,
T10N = T10 - 215,
T50N = T50 - 260,
T90N = T90 - 310,
Tx is the temperature (in C) at which x vol % of the sample has distilled off.
Despite its empirical nature, this equation has enjoyed enviable success in describing
the CN of various diesel fuels and their blends and is used extensively within the petroleum
industry. However, with the increasing trends toward low sulfur diesel fuels and the addition
of cetane improvers in the fuel, the results of this equation have become progressively less
satisfactory, and newer and better models are sought. In contrast to bulk property correlations,
many predictive models based on molecular composition have also been proposed to predict
CN. These include models based on NMR6-8 and FTIR9 spectra and the molecular
2
composition measured by a combination of liquid chromatography and GC-mass
spectroscopy (MS) techniques. Although composition-based, most of these models are linear
(and occasionally nonlinear) principal component regression models and do show
encouraging results in predicting CN. However, the lack of a mathematical structure for the
functional form and the assumption of linearity (in most cases) make such models inadequate
for extrapolation. Further, these models are typically refinery or application specific and
require frequent updating, thus restrictive in their application for any diesel fuel. In this work,
methods other than the ones mentioned above have been discussed to predict the cetane
number accurately. Most of these methods are statistical in nature, as a huge experimental
data has been used. Following are the methods.
A) Determination of cetane number of blend fuel from thermal properties
Based on the discussion made by K. Sivaramakrishnan and P. Ravikumar there exists a
relation between cetane number and properties of biodiesel so a correlation was developed
which is given as equation:
CN K1* K 2 * HV K 3 * FP K 4 *
Where,
K1, K2, K3, K4 are constants,
is kinematic viscosity (mm2/sec),
HV is heating value (MJ/kg),
FP is flash point (C),
is density (kg/l).
The above equation shows the relationship between the cetane number and the thermal
properties. Using this relation the cetane number for oils, biofuel, biodiesel blends were
calculated by K. Sivaramakrishnan1 and P. Ravikumar [2]. The average measured values of
CN obtained compare well with the predicated values using the generated equation. However
there are slight variations in rapeseed, sunflower and palm. This is due to high viscosity and
high calorific value of the vegetable oils. The predicted values, however, fall within the range
of measured values. The equation developed was used to predict the CN values of different
biodiesels consisting of different thermal properties.

3
Biodiesel Jatropha blends
Blend Measured Calculated Absolute Error Error %
B0 51 51.08 0.08 0.17
B20 51.2 51.15 0.05 0.09
B40 51.4 51.59 0.19 0.38
B60 51.9 51.99 0.09 0.17
B80 52.2 52.43 0.23 0.45
B100 53 53.1 0.1 0.2

Determination of cetane number of blend fuel from cetane number of parent fuels

As per the discussion by K. Sivaramakrishnan1 and P. Ravikumar [2], for biodiesel, the
nature of variation of cetane number with percent blend is linear in nature. So, as per the data
available for blends, a linear formula was formed. This formula is as follows:
CN (1 x) * CN 1 x * CN 2 ,
where,
x is the blend percent,
CN1 is the cetane number of diesel fuel,CN2 is the cetane number of bio-diesel fuel.
The predicted values using this linear relationship compare well with the measure value.

3. Correlation of ignition delay and cetane number of diesel-biodiesel


blend fuels:

4
Hardenberg and Hase equation:
An empirical formula, developed by Hardenberg and Hase for predicting the duration of the
ignition delay period in DI engines, has been shown to give good agreement with
experimental data over a wide range of engine conditions. This formula gives the ignition
delay (in crank angle degrees) in terms of charge temperature T (kelvins) and pressure p
(bars) during the delay (taken as TC conditions) as:-

1 1 21.2
id (CA) (0.36 0.22 * Sp) exp[ EA( ) * (( )^0.63))]
R * T 17190 p 12.4

where,
Sp, is the mean piston speed (meters per second),
R is the universal gas constant (8.3143 J/mol-K),
EA (joules per mole) is the apparent activation energy, and is given by:-
618840
EA
CN 25
where CN is the fuel cetane number.
The apparent activation energy decreases with increasing fuel cetane number. The delay in
milliseconds is given by:-
id (CA)
id (ms)
0.006 * N
where,
N is the engine speed in revolutions per minute.

5
4. Matlab Programs:
a) Thermal model for predicting cetane number of biodiesel (Karanja)
function CN = thermalmodel(nu,HV,FP,den)
% This program calculates the cetane number from thermal properties of the
% fuel(Karanja Fuel)
% Input 'nu' is kinematic viscosity of the fuel in mm2/sec
% Input 'HV' is heating vlue of the fuel in MJ/kg
% Input 'FP' is flashpoint of the fuel in C
% Input 'den' is specific density of fuel
% Output 'CN' is the predicted cetane number
A = [2.71 42.5 55 0.836; 4.01 41.5 65 0.849; 5.23 39.9 77 0.858; 6.72 38.7
88 0.862; 8.19 37.0 101 0.878];
b = [51; 51.70; 52.82; 53.15; 53.86];
y = A\b;
K1 = y(1,1);
K2 = y(2,1);
K3 = y(3,1);
K4 = y(4,1);
CN = K1*nu+K2*HV+K3*FP+K4*den;

b) Linear model for predicting cetane number (Karanja Fuel)


function CNBDF = linearmodel(CN1,CN2)
% This program calculates the cetane number of blend fuels of diesel
% bio-diesel
% Input 'CN1' is cetane number of diesel
% Input 'CN2' is cetane number of bio-diesel
% Output 'CNBDF' is cetane number of bio-diesel blend
Vf = [0:0.2:1];
CNBDF = CN1.*(1 - Vf) + CN2.*Vf;
CNBDM = [51 51.7 52.82 53.15 53.86 54.53];
plot(CNBDM,CNBDF,'*');
xlabel('measured cetane number');
ylabel('predicted cetane numbe');
x = CN1:CN2;
y = x;
hold on
plot(x,y);

6
measured cetane number vs predicted cetane number
55

54.5

54
predicted cetane numbe

53.5

53

52.5

52

51.5

51
51 51.5 52 52.5 53 53.5 54 54.5 55
measured cetane number

c) Hardenberg and Hase equation


function ID = ignition_delay(CN1,CN2,L,N,p,T)
% This program is used to calculate ignition delay of diesel-biodiesel
% blend fuels.
% Input 'PBc is percent blend'
%L = .1; % Input 'L is stroke length in metres'
%N = 4000; % Input 'N is rpm'
%p = 40; % Input 'p is charge pressure in atm'
%T = 400;% Input 'T is charge temperature in Kelvin'
%CN1 = 51; % Input 'CN1' is cetane number of diesel fuel
%CN2 = 53; % Input 'CN2' is cetane number of biodiesel fuel
R = 8.3143;
% Calcualtion of cetane number of diesel-biodiesel blend
Vf = [0:0.1:1];
CNBDF = CN1.*(1 - Vf) + CN2.*Vf;
% Calculation of activation energy
EA = 618840./(CNBDF + 25);
% Calcualtion of mean piston speed
Sp = 2*L*N/60;

% Calculation of ignition delay using Hardenberg and Hase method

7
ID_CA = (0.36+0.22*Sp)*exp(EA*(1/(R*T)-1/17190)*(21.2/(p-12.4))^0.63);
ID = ID_CA/(0.006*N);
figure(1);
plot(Vf,ID,'-r');
xlabel('Volume fraction');
ylabel('Ignition Delay(ms)');
title('Variation of ignition delay with blend percentage of
Biodiesel(Jatropha)');
figure(2);
plot(CNBDF,ID,'');
xlabel('cetane number');
ylabel('ignition delay(ms)');
title('Variation of ignition delay with ceatne number(Jatropha)');

Variation of ignition delay with blend percentage of Biodiesel(Jatropha)


0.665

0.66

0.655
Ignition Delay(ms)

0.65

0.645

0.64

0.635

0.63
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Volume fraction

8
Variation of ignition delay with cetane number(Jatropha)
0.665

0.66

0.655
ignition delay(ms)

0.65

0.645

0.64

0.635

0.63
51 51.2 51.4 51.6 51.8 52 52.2 52.4 52.6 52.8 53
cetane number

References

1. J. Hongbo a , H. Yanhui a & W. Huixin, A Model for Blending Properties of Diesel Oils and
Its Optimization Application, Petroleum Science and Technology
2. K. Sivaramakrishnan and P. Ravikumar, Determination of cetane number of biodiesel and
its influence on physical properties, published in ARPN Journal of Engineering and
Applied Sciences, Vol. 7, No. 2, February 2012
3. J.B.Heywood, Internal Combustion Engine Fundamentals
4. P.A.Laxminarayan, Yogesh V. Aghav, Modeling Diesel Combustion
5. P.K. Bose, Empirical Approach For Predicting The Cetane Number Of Biodiesel,
published in

9
Prediction of Ignition Delay and Cetane Numbers of blend of Diesel Fuel
with Kerosene and Terpentine

Abstract

In this report ignition delays of blend of 1.Diesel and kerosene and 2.diesel turpentine blend
are calculated by Hardenberg and Hase equations. Comparisons are made for ignition delays
versus cetane numbers. Also variations for the cetane numbers of various blends with respect
to their cetane numbers are seen. Plots for all the comparisons are plotted.

Cetane Number or CN is a measure of a fuel's ignition delay, the time period between the
start of injection and the first identifiable pressure increases during combustion of the fuel. In
a particular diesel engine, higher cetane fuels will have shorter ignition delay periods than
lower cetane fuels.
Generally, diesel engines operate well with a CN from 40 to 55. Fuels with higher cetane
number have shorter ignition delays, providing more time for the fuel combustion process to
be completed. Hence, higher speed diesel engines operate more effectively with higher cetane
number fuels.
Cetane number can be increased by ignition-accelerating additives like organic peroxides,
nitrates, nitrites and various sulphur compounds. Most important (commercially) is alkyl
nitrates about 0.5% by volume in a distillate fuel increases CN by 10.
Ignition delay is divided in to two parts as physical delay and chemical delay.

Ignition Delay

1.Physical Delay:

It is time required to for fuel atomization, vaporization, mixing with air and its temperature
rise.

2.Chemical Delay:

Chemical delay is time required for preflame reactions up to local ignition.

For delay period to be minimum, self ignition temperature of fuel is to be minimum. CN is


the scale for comparing the ignition delay for various fuels. More the cetane number lesser is
the ignition delay period and vice versa. Diesel fuels have cetane number in the range of 40 to
55. One of the solutions for increasing the cetane number of diesel is blending it with other
fuels having high cetane number. Among various fuels that can be blend to the diesel the
kerosene and turpentine are used in this report.

Hardenberg and Hase equation:

Hardenberg and Hase developed an empirical formula for calculating ignition delay period in
DI engines.

. .
(CA) = (0.36 + 0.22 Sp) exp[EA ( )( ]
.

Where,

Ignition delay in terms of crank angle rotations.


Sp- Mean piston speed in m/s
A- Apparent activation energy in J/mole

EA = +

CN - cetane number of fuel.

R Universal gas constant. (8.3143 J/mole K).


T Temperature during delay in Kelvin.
p- Pressure during delay in bar.

The delay in milliseconds is given by:-

CA
=
0.006

Where,

N- engine speed in revolutions per minute.


Diesel Blending

1.Turpentine Diesel Blending

Turpentine is a yellowish, opaque, sticky, volatile, combustible mixture of hydrocarbon


isomers obtained either from pine gum or pine wood. Turpentine has a minimum alpha
pinene content of 40 percent by weight. It consists chemically 5865% gamma pinene
along with beta pinene and other isometric terpenes.

Turpentine oil:-

It is obtained by distillation of resin which is obtained from trees generally pine trees. It is
combination of pinene (65-70)% and pinene (30)%. It is mainly used as thinner for paints. It
also has a medicinal use.

Properties of turpentine oil:-

1) colourless, characteristic odour and taste. 2) Soluble in 5 volume of alcohol. 3) When


perfectly pure, it exclusively consists of carbon and hydrogen.

Physical:-

1) boiling point-(153-175 )c. 2) flash point-35c. 3) viscosity-1.257mPas 4) specific gravity-


(.850-.865) 5) cp = 1720 (J/kgK)

Solubility:-Insoluble in water, soluble in benzene, chloroform, ether, carbon disulphide,


petroleum oils.

Advantages of turpentine oil over diesel:-

1) It is a renewable fuel, which is obtained from pine trees.

2) Self-ignition temperature of turpentine oil is near to that of diesel.

3) Boiling point is also almost same as that of diesel.

4) Calorific value of turpentine oil is higher than that of diesel.

5) It has same viscosity as that of diesel.

6) Compared with other bio fuels, turpentine has 11-15% higher calorific value.
Turpentine Diesel Blend Properties.

Items DF 50:50 60:40 70:30 TPOF

DF/TPOF DF/TPOF DF/TPOF

High caloric 45.89 41.26 42.12 43.21 37.28

value (MJ/kg)

Kinematic viscosity 3.6 5.30 4.80 4.70 3.89

(cSt @ 40 C)

Specic density 0.832 0.889 0.876 0.872 0.92

(g/cm2)

Cetane number 52 44 44 46 38

Diesel Kerosene Blend

Kerosene is the selected fuel for civic and military uses in jet-fuel turbine aircrafts. Therefore
it is an important task to assure the kerosene quality. From two multivariate studies several
tests have been considered as the most important ones in kerosene quality assurance initial
boiling point(IBP),fina lboiling point(FBP),distillation losses, freezing point, conductivity,
total acidity, olefins content, smoke point and water separation index. Due to refinery control
wishes, flash point(flash) and density(specific gravity) have been considered too.
Fuel Blend Results

SR CHEMICAL UNIT TEST


No PROPERTIES RESULTS
Sample Sample Sample Sample Sample
No.1 No.2 No.3 No.4 No.5
Product HSD HSD HSD HSD HSD75%
100% 90%+ 85%+ 80%+ + SKO
SKO SKO 15% SKO 25%
10% 20%
1 Density@15degC kg/m3 827.1 824.5 823.1 821.1 819.7
2 Calculated 54.5 54.1 53.1 52.6 51.8
Cetane Index

MATLAB PROGRAMS

Matlab Program for calculation of ignition delay of kerosene diesel blend.

function IDK=idk(L,N,T,p)

data1= xlsread('Book1.xlsx');
CN=data1(: ,1);

%L = .1; % Input 'L is stroke length in metres'


%N = 3000; % Input 'N is rpm'
%p = 40; % Input 'p is charge pressure in atm'
%T = 450;% Input 'T is charge temperature in Kelvin'
%CN = % Input 'CN' is cetane number of diesel fuel

R = 8.3143;
%R=Universal Gas constant in KJ/kg
% Calculation of activation energy
EA = 618840./(CN + 25);
% Calculation of mean piston speed
Sp = 2*L*N/60;

% Calculation of ignition delay using Hardenberg and Hase method


ID_CA = (0.36+0.22*Sp)*exp(EA*(1/(R*T)-1/17190)*(21.2/(p-12.4))^0.63);
IDK = ID_CA/(0.006*N);
fprintf('\n The ignition delay in millisecond is: %0.8f',IDK);
plot(CN,IDK,'-.or')
xlabel('cetane number');
ylabel('ignition delay(ms)');
grid on;
hold on;

2. Cetane Number and Blend percentage curve

Matlab Program

data1= xlsread('Book1.xlsx');
pbc=data1(: ,2);
CN=data1(: ,1);
plot(CN,pbc,'-.or')
ylabel('cetane number');
xlabel('kerosene percent blend');
grid on;

Curve Fit Equation estimation

function kerosene_cetane(x)

p1=-0.003857;

p2 =-0.01386;
p3 =54.51;

CN = p1*x^2 + p2*x + p3;


fprintf('\n cetane number is: %0.8f',CN);
plot(x,CN)

Diesel Turpentine Blend

Matlab Program

function idt=turpentine_ameya(L,N,T,p)

data1= xlsread('Book2_T.xlsx');
CN=data1(: ,1);

%L = .1; % Input 'L is stroke length in metres'


%N = 3000; % Input 'N is rpm'
%p = 40; % Input 'p is charge pressure in atm'
%T = 450;% Input 'T is charge temperature in Kelvin'
R = 8.3143;
%R=Universal Gas constant in KJ/kg

% Calculation of activation energy


EA = 618840./(CN + 25);
% Calculation of mean piston speed
Sp = 2*L*N/60;

% Calculation of ignition delay using Hardenberg and Hase method


IDP= (0.36+0.22*Sp)*exp(EA*(1/(R*T)-1/17190)*(21.2/(p-12.4))^0.63);
idt = IDP/(0.006*N);
fprintf('\n The ignition delay in millisecond is: %0.8f',idt);
plot(CN,idt,'--ob')
xlabel('cetane number');
ylabel('ignition delay(ms)');
grid on;
hold on;
2.Percent Blends vs Cetane number

Matlab Program

data1= xlsread('Book2_T.xlsx');
pbc=data1(: ,2);
CN=data1(: ,1);
plot(pbc,CN,'--ob')
ylabel('cetane number');
xlabel('turpentine percent blend');
grid on;
Results and Conclusion

1.The curves shows that ignition delay for diesel kerosene and diesel turpentine blends
decreases with increase in cetane number.

2.The cetane number values for both the fuels decrease with increase in blend percentage of
other fuel

References

1.R.S.Bharj and Sundeep Katiyar , CHEMICAL ANALYSIS OF BLENDING OF DIESEL


AND KEROSENE FUELS USED IN SMALL CI ENGINE;
2nd International Conference on Role of Technology in Nation Building (ICRTNB-2013)

2. B. Prem Anand, C.G. Saravanan; Performance and exhaust emission of turpentine oil
powered direct injection diesel engine in journal of Renewable Energy 35 (2010) 11791184.
1

Determination of Ignition Delay with Cetane number of Diesel-Ethanol-


Biodiesel Blend
1) Introduction

The search for alternative fuels for currently used fossil fuels have led to many different fuels
that can be obtained from renewable sources or can be blended with the currently available fuels.
One such fuel is bioethanol. Due to the availability in large volume, among all the renewable fuel
sources bioethanol can be a good option especially because the second-generation production
process of bioethanol is going to be available very soon in the coming years. Bioethanol is
produced from various feedstocks like sugarcane, corn, beet, molasses, cassava root, barley
sugar, starch, cellulose, etc. In addition, ethanol can be produced from reacting ethene and steam
(As this process requires a lot of energy, a major portion of the world s ethanol is produced from
renewable sources through fermentation. Thus the bioethanol is mostly named as simply
ethanol.). Although it is regarded primarily as a substitute fuel for SI or spark ignition engines
but it also has potential uses for CI engines. During the 1980s the blending of ethanol with fossil
diesel fuel was a topic of research and it was investigated that this ethanoldiesel blend, also
known as diesohol, is technically usable in the then existing diesel engines without any major
modifications. The main advantage of ethanol is its oxygen content which is about 34% by
weight. Beside this oxygen content the use of ethanol with diesel fuel has many more advantages
as well as some major disadvantages. Taking all these factors under consideration the usage of
ethanol with diesel fuel in the diesel engines seems attractive.
But, the Cetane number of this blend becomes lower compared to diesel fuel. The addition of
10 %(v/v) of ethanol decreases cetane number by approximately 30%.

2) Cetane Number

Cetane number (CN) of a CI engine fuel can be defined as the measurement of the
combustion quality of a fuel during compression ignition. The cetane number and cetane index is
in proportion to the density value. The measurements those determine the overall fuel quality,
cetane number (CN) is a significant one among them. The requirements of a minimum cetane
number depend on the engine design, size, nature of speed, and load variations, as well as
starting and atmospheric conditions. Using fuels of cetane number higher than the requirement
does not materially increase the performance of the engine. Accordingly, to ensure maximum
fuel availability the specified cetane number should be up to a certain limit. Again using fuels
with low cetane number than the actual requirements could cause rough operation of the engine.
Problem like difficult starting, especially in the cold weather or at high altitudes are faced by
using this type of fuels. Formation of lube oil sludge is also accelerated by using this type of fuel.
Besides several low cetane number fuels are the cause of larger engine deposits which results in
extra smoke than usual, increased exhaust emissions and greater engine wear. Based on the
cetane number of the constituents and the mass composition of the blend, the cetane number of
2

that specific blend is assessed. The lower cetane number means longer ignition delays, allowing
more time for the fuel to vaporize before combustion starts. Initial burn rates are higher causing
more heat release at constant volume, which is a more efficient conversion process of heat to
work.
Nevertheless it is preferable to add an ignition improver like biodiesel to raise the cetane
number of dieselethanol blends so that they fall within an acceptable range equivalent to that
expected of No. 2 diesel fuel.

As specified by the ASTM Standard D 975-02 the minimum cetane number for No. 2 diesel
should be 40. Typically, the cetane number for No. 2 diesel fuels vary from 45 to 50. With the
opposite relationship of cetane number and octane number, ethanol displays a low cetane rating.
Hence as the ethanol concentrations in the diesel increases, the cetane number of the blend
decreases proportionately. The lesser the cetane number the shoddier will be the ignition
property. Engine performance, its start-up and the combustion control are affected by the cetane
number of the fuel being used.
It was projected that the cetane number of ethanol in between 5 and 15. Due to the very low
cetane number of the ethanol, the cetane number of the dieselbiodiesel ethanol blend
decreases. However, due to the high cetane number, biodiesel can recuperate this property, thus
the fuel blend can achieve the cetane number requirement for diesel which is 51 CN.

By experiments, it was found that decreasing of cetane number of the diesohol blend with the
increasing amount of ethanol in the blend, which was due to the low cetane number of ethanol
which is roughly 58. They also reported the same results as the previous study. They found that
the sample comprising of 80% diesel, 15% biodiesel and 5% ethanol has a maximum cetane
number which they measured to be 48.66.

Zldy[1] studied different ternary blends for measuring cetane number. The investigators here
also concluded that, there is a decrease in the cetane number of the blend with the addition of
ethanol and increase again with the addition of biodiesel, which matches the other described
results above. They developed a multi-linear model, which can be applied for modeling the
compensation factor of adding ethanol and biodiesel to diesel fuel. The model rose to be the like
as displayed below for the dependency of cetane numbers on the amount of ethanol and
biodiesel. The attained formula shows the dependency on the components:

CNebdg = CNd - 0.59 ne + 0.55 nbd

where, CNebdg is the calculated cetane number of the blend; CNd is the cetane number of diesel
oil; ne is ethanol ratio in the blend [v/v %]; nbd is biodiesel ratio in the blend [v/v%].
Due to the ethanol addition a decrease of 0.6 in cetane number is well-compensated by adding
biodiesel, which again increases the cetane number by an amount of 0.55. They suggested that if
it is desired to hold the cetane number of the blend at its original value even after adding ethanol
and biodiesel then these two components should be blended approximately at a 1:1 ratio with the
3

diesel fuel.

Park et al.[2] concluded that, increasing the biodiesel amount in the blend increases the blend
cetane number as the biodiesel fuel has a greater cetane number compared to both diesel fuel and
ethanol. Normally, the cetane number of the used biodiesel is associated with the varieties of
feed-stocks used in its making. In addition, the long fatty acid carbon chains and more saturated
molecules results in a high cetane number.

3) Hardenberg and Hase method to determine ignition delay

An empirical formula, developed by Hardenberg and Hase for predicting the duration of the
ignition delay period in DI engines, has been shown to give good agreement with experimental
data over a wide range of engine conditions (see Fig. 10-3). This formula gives the ignition delay
(in crank angle degrees) in terms of charge temperature T (Kelvin) and pressure p (bars) during
the delay (taken as TC conditions) as:-

. .
_ (CA) = (0.36 + 0.22 Sp) exp[EA ( )( ]
.

where Sp, is the mean piston speed (meters per second) and R is the universal gas constant
(8.3143 J/mol-K). EA (joules per mole) is the apparent activation energy, and is given by:-

EA = +

where CN is the fuel cetane number. The apparent activation energy decreases with increasing
fuel cetane number.

The delay in milliseconds is given by:-

CA
=
0.006

where N, engine speed, is in revolutions per minute.

4) Matlab Programs
4

1) Calculation of ignition delay by Hardenberg and Hase method:-


function t = ignitiondelaycn(CNd,ne,nbd,L,N,T,P)

% CNd = 52; Input 'CNd is cetane number of diesel'


% ne = 0.05; Input 'ne is ethanol ratio in the blend'
% nbd = 0.1; Input 'nbd is biodiesel ratio in the blend'
% L = 0.1; Input 'L is stroke length in metres'
% N = 4000; Input 'N is speed in rpm'
% P = 45; Input 'P is charge pressure in atm'
% T = 500; Input 'T is charge temperature in Kelvin'

R = 8.314;
%------Calculation of mean piston speed--------
Sp = 2*L*(N/60);

%------Calculation of cetane number of diesel ethanol biodiesel blend------


CN = CNd - 0.59*ne + 0.55*nbd;

%------Calculation of activation energy--------


Ea = 618840 / (CN + 25);

%------Calculation of ignition delay using hardenberg and Hase method------


tca = (0.36+0.22*Sp)*exp(Ea*(1/(R*T)-1/17190)*(21.2/(P-12.4))^0.63);
t = tca / (0.006 * N);

-----------------------------------------------------------------------------
This program is used to determine the ignition delay in milliseconds
when the input is in terms of blend percentage.
The inputs required are the fraction of ethanol and biodiesel in the
blend along with the stroke length of the piston, engine speed,
temperature and pressure of the charge during the delay.
The sample values of the input are mentioned in the program.
When the function named ignitiondelaycn is called and the necessary
parameters are put into it, the user gets the ignition delay in
milliseconds.
-----------------------------------------------------------------------------

2) Variation of ignition delay with cetane number


function t = ignitiondelaycn1(CNd,L,N,T,P)

% CNd = 51; Input 'CNd is cetane number of diesel'


% ne = 0.05; Input 'ne is ethanol ratio in the blend'
% nbd = 0.1; Input 'nbd is biodiesel ratio in the blend'
% L = 0.1; Input 'L is stroke length in metres'
% N = 4000; Input 'N is speed in rpm'
% P = 45; Input 'P is charge pressure in atm'
% T = 500; Input 'T is charge temperature in Kelvin'
R = 8.314;
%------Calculation of mean piston speed--------
5

Sp = 2*L*(N/60);

%------Calculation of cetane number of diesel ethanol biodiesel blend------


ne = 0:2:20;
nbd = 0:2:20;
CN = CNd - 0.59*ne + 0.55*nbd;

%------Calculation of activation energy--------


Ea = 618840 ./ (CN + 25);

%------Calculation of ignition delay using hardenberg and Hase method------


tca = (0.36+0.22*Sp)*exp(Ea*(1/(R*T)-1/17190)*(21.2/(P-12.4))^0.63);
t = tca / (0.006 * N);

figure(1);
plot(CN,t);
xlabel('cetane number');
ylabel('ignition delay');

-----------------------------------------------------------------------------

This program gives the variation of ignition delay with cetane number of the blend.
The inputs required are the cetane number of the diesel, stroke length of piston, engine
speed, temperature and pressure of the charge during the delay.
When the function ignitiondelaycn1 is called and the necessary parameters are input by
the user, he can see that with increase in cetane number the ignition delay decreases.
-----------------------------------------------------------------------------
6

3) Variation of cetane number with volume fraction of ethanol

function CNbd = linearbd(CNd)

% CNd = 52; Input cetane number of diesel

%-----Volume Fraction of Biodiesel = 5%-------


nbd = 0.05;

ne = 0:0.05:0.4;

CNbd = CNd - 0.59 * ne + 0.55 * nbd;

plot(ne,CNbd);
xlabel('Volume fraction of ethanol');
ylabel('Cetane Number');
title('Variation of cetane number with blend % of ethanol');
hold on; grid on;

%-----Volume Fraction of Biodiesel = 10%--------


nbd1 = 0.1;

ne = 0:0.05:0.4;

CNbd = CNd - 0.59 * ne + 0.55 * nbd1;

plot(ne,CNbd,'m');
xlabel('Volume fraction of ethanol');
ylabel('Cetane Number');
title('Variation of cetane number with blend % of ethanol');

%-----Volume Fraction of Biodiesel = 20%-----


nbd2 = 0.2;

ne = 0:0.05:0.4;

CNbd = CNd - 0.59 * ne + 0.55 * nbd2;

plot(ne,CNbd,'k');
xlabel('Volume fraction of ethanol');
ylabel('Cetane Number');
legend('Biodiesel = 5%','Biodiesel = 10%','Biodiesel = 20%');
title('Variation of cetane number with blend % of ethanol');
7

-----------------------------------------------------------------------------

This program gives us the variation of cetane number with volume fraction of ethanol for
constant volume fraction of biodiesel.
The input required is only the cetane number of diesel.
When the function linearbd is called and the necessary parameter is put into it, we get a
graph showing that the cetane number decreases linearly with increase in percentage of
ethanol for a given percentage of biodiesel.
Also, as the percentage of biodiesel in the blend is increased, the effective cetane number
increases.
-----------------------------------------------------------------------------
8

4) Variation of cetane number with volume fraction of biodiesel


function CNet = linearet(CNd)

% CNd = 52; Input cetane number of diesel

%-----Volume Fraction of Ethanol = 5%-----


ne = 0.05;

nbd = 0:0.05:0.4;

CNet = CNd - 0.59 * ne + 0.55 * nbd;

plot(nbd,CNet);
xlabel('Volume fraction of biodiesel');
ylabel('Cetane Number');
title('Variation of cetane number with blend % of biodiesel');
hold on;

%-----Volume Fraction of Ethanol = 10%-----


ne1 = 0.1;

nbd = 0:0.05:0.4;

CNet = CNd - 0.59 * ne1 + 0.55 * nbd;

plot(nbd,CNet,'k');
xlabel('Volume fraction of biodiesel');
ylabel('Cetane Number');
title('Variation of cetane number with blend % of biodiesel');
hold on; grid on;

%-----Volume Fraction of Ethanol = 20%------


ne2 = 0.2;
nbd = 0:0.05:0.4;

CNet = CNd - 0.59 * ne2 + 0.55 * nbd;

plot(nbd,CNet,'m');
xlabel('Volume fraction of biodiesel');
ylabel('Cetane Number');
legend('Ethanol = 5%','Ethanol = 10%','Ethanol = 20%');
title('Variation of cetane number with blend % of biodiesel');
9

-----------------------------------------------------------------------------

This program gives us the variation of cetane number with volume fraction of biodiesel
for constant volume fraction of ethanol.
The input required is only the cetane number of diesel.
When the function linearet is called and the necessary parameter is put into it, we get a
graph showing that the cetane number increases linearly with increase in percentage of
biodiesel for a given percentage of ethanol. Also, as the percentage of ethanol in the
blend is increased, the effective cetane number decreases.
-----------------------------------------------------------------------------

5) REFERENCES
[1] Zldy M; Ethanolbiodieseldiesel blends as a diesel extender option on compression
ignition engines, Transport 2011; 26:3039
[2] Park SH, Cha J, Lee CS. Impact of biodiesel in bioethanol blended diesel on the engine
performance and emissions characteristics in compression ignition engine, Appl Energy
2012; 99:33443
[3] S.A.Shahir n, H.H.Masjuki, M.A.Kalam, A.Imran, I.M. Rizwanul Fattah, A. Sanjid;
Feasibility of dieselbiodieselethanol/bioethanol blend as existing CI engine fuel: An
assessment of properties, material compatibility, safety and combustion; Centre for Energy
Sciences, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia.
[4] Istvn Barabs and Ioan-Adrian Todoru Technical University of Cluj-Napoca Romania,
Utilization of Biodiesel-Diesel-Ethanol Blends in CI Engine.
PREDICTION OF IGNITION DELAY IN CI ENGINES FOR DIFFERENT
BLENDS OF DEE AND DIESEL BY USING ARRHENIUS,
HARDENBERG-HASE AND SHIPINSKI EQUATION

Abstract:

In this paper ignition delay of blend of Diesel and Diethyl-ether are calculated by
various equations given by different authors. Ignition delays calculated by different
methods are then compared to the actual data available from experimental data. Trend of
blend percentage versus cetane number are plotted to know effect of blending on cetane
number and ultimately on ignition delay.

Introduction:

The ignition delay is time between start of injection to the point where the P-
curve separates from the pure air compression curve. It is very important to study delay
period and factors affecting delay period as it has great influence on engine design and
performance. When delay period is high more quantity of fuel is injected into the
combustion chamber and therefore there is high rate of pressure rise in uncontrolled
phase of combustion which leads to knocking, with keeping this phenomenon in mind,
engine designer should always try to keep delay period as minimum as possible. On the
other hand, when delay period is too minimum, then there will not be sufficient time for
dispersing the fuel droplets in the chamber which results in smoke and high fuel
consumption. By considering above two things optimum value of delay period should be
selected for smoother engine operation. In actual practice, ignition delay period is more
than desired so designer is always trying to reduce it. Delay period is measured either in
terms of absolute time or crank angle rotations. When engine speed increases time
available for heat transfer from compressed air to cylinder wall decreases therefore both
temperature and pressure of compressed air increases therefore delay period in terms of
milliseconds decreases but delay period in terms of crank angle rotations increases with
increase in speed. As delay period in terms of crank angle rotations increases more fuel is
injected as fuel pump is geared to the engine. This is one of the major disadvantages for
making very high speed diesel engine

Ignition delay is divided in to two parts as physical delay and chemical delay.

1. Physical Delay:

It is time required to for fuel atomization, vaporization, mixing with air and its
temperature rise.

2. Chemical Delay:

Chemical delay is time required for preflame reactions up to local ignition.

For delay period to be minimum, self ignition temperature of fuel is to be minimum. CN


is the scale for comparing the ignition delay for various fuels. More the cetane number
lesser is the ignition delay period and vice versa. Diesel fuels have cetane number in the
range of 40 to 55. One of the solutions for increasing the cetane number of diesel is
blending it with other fuels having high cetane number. Among various fuels that can be
blend to the diesel the DEE is becoming more common now days.

Diethyl ether (DEE):

Diethyl ether has chemical formula as C2H5-O-C2H5. It contains less number of carbon-
to -carbon bonds therefore when it is blended with diesel the concentration of solid
carbon particulate during combustion decreases. DEE has a cetane number greater than
125. This property of DEE is very important for reducing the delay period. DEE has auto
ignition temperature of 1800 C near to self ignition temperature of diesel (2500C) which is
favorable condition for blending DEE with diesel.

DEE Diesel Blend Properties:

Sr. No Fuel blend DENSITY@15OC CN


(Kg/m3)
1 D100 836 52
2 DE2D 833 53.46
3 DE5D 829 55.65
4 DE8D 826 57.84
5 DE10D 823 59.3
6 DE15D 817 62.95
7 DE20D 811 66.6
8 DE25D 805 70.25
9 DE100 713 125

1. Arrhenius Equation:

Arrhenius equation for calculating ignition delay is as below:


= exp

Where,
= Ignition delay in milliseconds.

= Apparent activation energy for the fuel auto-ignition process.

EA =
+

= Universal gas constant.

and = Constants dependent on the fuel and are given by wolfer in table below.

P = Pressure at top centre position of piston in atm.

T = Temperature at top centre position of piston in Kelvin.


Investigator Conditions Parameters

Appara Fuel , atm , K


tus
Wolfer Consta Fuel with 8-48 590- 1.19 0.44
nt cetane 782
volume number >
bomb 50

2. Hardenberg and Hase equation:

Hardenberg and Hase developed an empirical formula for calculating


ignition delay period in DI engines.

. .
(CA) = (0.36 + 0.22 Sp) exp[EA ( )( ]
.

Where,
Ignition delay in terms of crank angle rotations.
Sp- Mean piston speed in m/s
A- Apparent activation energy in J/mole

EA = +

CN - cetane number of fuel.

R Universal gas constant. (8.3143 J/mole K).


T Temperature during delay in Kelvin.
p- Pressure during delay in bar.

The delay in milliseconds is given by:-


CA
=
. 6

Where,

N- engine speed in revolutions per minute.

Values for T and p can be estimated using a polytropic model for the compression
process:


= =

Where,

n - polytropic exponent.

rc - compression ratio.

Subscript i denotes intake manifold conditions. Values of the polytropic exponent


are given in Fig. below for a direct-injection diesel under warm and cold engine
operating conditions.
3. Shipinski equation:

.
6 .8 6 4644 4
= + exp ( )( )
.

Where,

Ignition delay in terms crank angle rotation.

Injection duration in terms of crank angle rotations.

. N Engine RPM.

P Average pressure in atm.


t Average temperature in Kelvin.

Programmes:

1. Programme to plot a graph of cetane vs ignition delay in terms of crank angle rotations.
function ID_CA=ignitiondelay_DEE(L,N,T,p)

data1= xlsread('ashish.xlsx');
CN=data1(: ,2);

%L = .1; % Input 'L is stroke length in metres'


%N = 3000; % Input 'N is rpm'
%p = 40; % Input 'p is charge pressure in atm'
%T = 450;% Input 'T is charge temperature in Kelvin'
%CN = % Input 'CN' is cetane number of diesel fuel

R = 8.3143;
%R=Universal Gas constant in KJ/kg

% Calculation of activation energy


EA = 618840./(CN + 25);
% Calculation of mean piston speed
Sp = 2*L*N/60;

% Calculation of ignition delay using Hardenberg and Hase method


ID_CA = (0.36+0.22*Sp)*exp(EA*(1/(R*T)-1/17190)*(21.2/(p-12.4))^0.63);
%ID_DEE = ID_CA/(0.006*N);
fprintf('\n The ignition delay in terms of crank angle is: %0.8f',ID_CA);
%plot(CN,ID_DEE,'-.or')
plot(CN,ID_CA)
xlabel('cetane number');
ylabel('ignition delay(ca)');
grid on;
2. Programme to plot blend percentage vs cetane number

data1= xlsread('ashish.xlsx');
pbc=data1(: ,1);
CN=data1(: ,2);
plot(CN,pbc,'-.or')
ylabel('cetane number');
xlabel('DEE percent blend');
grid on;

3.Programme to compare ignition delay of various methods


% program for comparison of ignition delays by various methods///
CN = [52, 53.46, 55.65, 57.84, 59.3, 62.95, 66.6, 70.25];
Tid_theta = [18, 20 , 18, 17.5, 17.3, 16, 15.3, 15]
plot(CN,Tid_theta,'b'),grid;
hold on,
Ti = 300;
Pi = 1;
rc = 18;
n = 1.3;
R = 8.314;
T = Ti.*rc^(n-1);
P = Pi.*rc^n;
N = 1500;
L = 0.234;
% calculation of mean piston speed///
Sp = (2*L*N)./60;
EA = 618840./(CN+25);
A = 0.44;
m = 1.19;
% Arrhenius equation///
Tid_ms = A.*P.^m.*exp(EA./(R.*T));
Tid_theta = Tid_ms.*0.006.*N./100;
plot(CN,Tid_theta,'y');
hold on;
% Hardenberg and Hase method///
Tid_theta = ((0.36+022.*Sp).*exp(EA.*((1./(R.*T))-(1./17190)).*(21.2./(P-12.4)).^0.63))./100;
plot(CN,Tid_theta,'g');
hold on;
% shipinski method///
thetainj = 23;
Tid_theta = (thetainj+(6.*N.*0.8216)./(1000.*P^0.386).*exp(4644/T).*(40./CN).^0.69)./100;
plot(CN,Tid_theta,'r');
xlabel('Cetane number of blend of DEE and Diesel');
ylabel('Ignition delay in crank rotations');
legend('Actual','arrhenius','Hardenberg & Hase','Shipinski');
Results & Conclusions:

1. With increasing blend percentage of DEE with diesel ignition delay in terms of crank
rotations is initially increasing but later it decrease.
2. Arrhenius equation is giving close predictions to the actual data of ignition delay.
3. Relation between Cetane number and blend percentage is linear.

References:

1. Experimental investigation of CI engine combustion, performance and emissions in


DEE kerosenediesel blends of high DEE concentration by S.S. Thipse and K.R. Patil.
2. Internal combustion engine fundamentals by John B. Heywood.
The Heat Release Rate

The heat release rate describes the temporal course of energy release in the combustion
chamber. The integral of the heat release rate is designated as the total heat release or the burn
function. In order to model combustion in the context of process calculation we use various
approaches and mathematical modelling, which all have the goal of describing the actual heat
release via combustion as exactly as possible as so-called substitute heat release rates.
Another possibility for describing heat release are so-called phenomenological models, that
pre-calculate the heat release rate, starting from the injection path of the diesel engine for
example in recent times, purely numerical methods are being used because of the increasing
complexity of combustion processes (e.g. direct injection with stratified charge in the SI
engine, common-rail injection in the diesel engine with multiple injection). In this case we are
dealing with so-called neural networks, which have to be trained with results from measured
operating points and can be contingent on a number of parameters.

7.1.7.1 The Vibe Substitute Heat Release Rate

Proceeding from the triangular combustion, Vibe (1970) provided the following relation for
the total heat release rate with the help of reaction kinetic consideration

m 1
SOC
Q f ( ) a
CD


1 e mit SOC ( SOC CD ) . (7.64)
Q f ,total

The entire amount of released energy is thereby to be calculated from the product of the fuel
mass brought into the combustion chamber and the lower heating value
Q f ,total m fuel I hv . (7.65)

Furthermore, QSOC signifies the start of combustion and CD the combustion duration. The
so-called Vibe shape parameter is designated with m. At the end of combustion, i.e. at
EOC , a certain percentage conv,total of the total energy added with the fuel must be
converted (degree of conversion), for which is valid
Q f ( )
conv,total 1 e a . (7.66)
Q f ,total
EOC

From this follows for the factor a the relation


a ln(1 conv,total );

from which we obtain the following numerical values


If we derive the total heat release rate from the degree of crank angle, we obtain the following
for the heat release rate:
m 1
m SOC
dQ f SOC a
CD


Q f ,totala(m 1) e . (7.67)
d CD
Figure 7.12 shows heat release rates for various Vibe form factors. One should note here that
Vibe form factors smaller than 0 are also possible. In order to adjust a real combustion by
means of a Vibe substitute heat release rate, there are various methods for determining the
three Vibe parameters start of combustion, combustion duration, and shape parameter. The
parameters can either be adjusted visually or can be determined using mathematical methods
(e.g. that of Rel. crank the least square). It is however important that important process data
such as peak pressure, indicated mean pressure and the exhaust gas temperature are calculated
with the substitute heat release rate in agreement with reality. The determination of the Vibe
parameters is de rigueur in most pressure analysis programs.
Fig. 7.12 Vibe substitute heat release rate for different Vibe shape coefficients

7.1.7.2 The Substitute Heat Release Rate as a Double Vibe Function

In engines or operating points with a clearly marked premixed combustion such as in diesel
engines, reproduction with a simple Vibe substitute heat release rate is usually too imprecise.
We thus often replace the simple Vibe substitute heat release rate with the superimposition of
two Vibe functions, the so-called double Vibe function. If one describes an actual heat release
rate with a double Vibe function, a subdivision of the energy shares of both functions is also
necessary besides the two simple Vibe heat release rates. For a double Vibe heat release rate
is valid
m1 1
m1 SOC,1
dQ f ,1 SOC,1 a CD ,1

Q f ,1a(m1 1) e

. (7.68)
d CD ,1
mit SOC,1 ( SOC,1 CD,1 ),
m2 1
m2 SOC, 2
dQ f , 2 SOC, 2 a



Q f , 2 a(m2 1) e

CD , 2
. (7.69)
d CD, 2
mit SOC, 2 ( SOC, 2 CD, 2 ),

Q f ,1 xQ f , ges and Q f , 2 (1 x)Q f , ges , (7.70)

dQ f dQ f ,1 dQ f , 2
. (7.71)
d d d

The diagram on the left in Fig. 7.13 shows the reproduction of the heat release rate in the
rated performance point of a high speed high performance diesel engine via a double Vibe
substitute heat release rate. One can clearly see that the double Vibe function cannot exactly
reproduce combustion lasting until the exhaust phase. This has to do with the mathematical
form of this substitute heat release rate. The diffusion phase is described by the central
exponential term of the double Vibe function. This exponential term with its asymptotic
course towards burning matter can under certain circumstances still not exactly describe large
energy release rates at the burning location in the concrete case, which is why on occasion a
threefold Vibe function is used instead of the double Vibe function. The exact reproduction of
premixed combustion is an important prerequisite for a good agreement of the working cycle
calculation with measurements, which is why the adjustment and weighing of the first Vibe
function should receive careful consideration. One problem in the double Vibe function is as
in the case of the simple Vibe function as well the exact description of the burn-out phase,
which is, especially in charged engines, responsible for the exhaust gas temperature level.
Good indicator for this is the total heat release rate, in which deviations are most clearly
visible. The method of the least square has proved to be the most suitable adjustment routine
in the case of the double Vibe function as well.
Fig. 7.13 The double-Vibe function and polygon-hyperbola heat release rate

7.1.7.3 The Polygon-Hyperbola Substitute Heat Release Rate

In view of the calculation of NOx formation by means of multi-zone models, a further


substitute heat release rate, the polygon-hyperbola substitute heat release rate, was suggested
by Schreiner (1993), which is represented in Fig. 7.13 on the right. It consists of a polygonal
port (1-4-5) and a hyperbola connected to it (5-6). A triangle (1-2-3) is superimposed over the
polygonal port, which serves in the description of premixed combustion. For the mathematical
description of this polygon-hyperbola substitute heat release rate, we need nine parameters.

Premixed phase:

Q f , pre ( 1 )
y 2* with 1 2 (7.72)
d ( 2 1 )

Q f , pre ( 3 )
y 2* with 2 3 (7.73)
d ( 3 2 )

Diffusion phase:

Q f ,diff ( 1 )
y4 with 1 4 (7.74)
d ( 4 1 )

Q f ,diff
y4 with 4 5 (7.74)
d
Q f ,diff ( y 4 y6 ) ( y 4 y6 )
h3 h1 ( 1 ) h2 y 4 ( 1 ) h2
d ( 1 )
h2
( 5 1 ) ( 6 1 )
h2 h2

1 6
( 5 1 )

mit 5 6 (7.76)

for the entire combustion process

dQ f ,total dQ f , pre dQ f ,diff


. (7.77)
d d d

is valid.
The quantity y 2* sets the height of the peak in premixed combustion. The threehyperbolic
parameters h1, h2 and h3 result from the conditions that the hyperbola goesthrough points 5
and 6 and that the integral under the substitute heat release rate hasthe value of 1. h2 thereby
fixes the path of the hyperbola. The maximum height ofthe substitute heat release rate in the
diffusion phase y4 results from the conditionthat the centre of the substitute heat release rate
agree with the experimentallydetermined heat release rate. The quantity y6 describes the
conversion at the end ofcombustion.The agreement in Fig. 7.13 (right) is very good, such that
there practically nodisparity arises in the results of the actual working cycle calculation
between thepoint-for-point provision and the approximation of experimental data with this
Substitute heat release rate.Further details and an extensive analysis of the polygon-hyperbola
substituteheat release rate can be found in Schreiner (1993).
Matlab program

%polygone-hyperbo5la heat release rate%

clc
closeall
x1=2.8121;%CA of start of pre-combustion phase(in polygon)
x2=6.1867;%CA of peak of pre-combustion phase(in polygon)
% x3=8.9989;%CA of end of triangular diffuse-combustion phase(in polygon)
x3=7.3149;%CA of intersection of triangular and diffuse-combustion phase(in
polygon)
x4=11.2486;%CA of peak of diffuse-combustion phase(in polygon)
x5=13.1856;%CA of peak of diffuse-combustion phase(in polygon)
x6=90;%CA of end of diffuse-combustion phase(in hyperbola)
y1=0.41;%height of peak in premixed combustion
y4=0.46;%maximum height of substitute heat release rate in the diffuse
phase

x=linspace(x1, x2, 100);%curve divided in to 100 parts from x1 to x2


dq1=y1.*(x-x1)./(x2-x1);%first heat release rate curve equation
plot(x,dq1)
holdon

% x=linspace(x2, x3, 100);


% dq2=y1.*(x3-x)./(x3-x2);
% plot(x,dq2)
% % line([x2,y1],[x3,0.2455])
% hold on

% x=[x2:100:x3];
x=linspace(x2, x3, 100);%curve divided in to 100 parts from x2 to x3
y=-0.1458.*x+1.3121;%second heat release rate curve equation
plot(x,y)
holdon

% x=linspace(x1, x4, 150);%curve divided in to 100 parts from x1 to x4


x=linspace(x3, x4, 150);%curve divided in to 100 parts from x3 to x4
dq3=y4.*(x-x1)./(x4-x1);%third heat release rate curve equation
plot(x,dq3)
holdon
x=linspace(x4, x5, 100);%curve divided in to 100 parts from x4 to x5
dq4=y4;%forth heat release rate curve equation
plot(x,dq4)
holdon

x=linspace(x5, x6, 300);%curve divided in to 100 parts from x5 to x6


y6=0.01;%conversion at the end of combustion
h2=-1;%value is adjusted thereby it fixes the path of hyperbola
h1=(y4-y6)./((x5-x1).^h2-(x6-x1).^h2);%hyperbolic parameter
h3=y4-(y4-y6)./(1-((x6-x1)./(x5-x1)).^h2);%hyperbolic parameter
dq5=h3+h1.*(x-x1).^h2;%fifth heat release rate curve equation
plot(x,dq5)
holdon

gridon
title('heat release rate vs crank angle')
xlabel('crank angle(CA)')
ylabel('heat release rate(kJ/CA)')
Conclusion

Polygon-hyperbola heat release rate method gives more accurate results


with respect to experimental results of actual heat release rate to that of double
vibe heat release rate method
Calculation for Heat Release Rate for Indirect-injection Engine

In IDI diesel engines, the pressure in each of the two chambers, main and auxiliary, are not same
during combustion process. Since combustion starts in the auxiliary or prechamber, the fuel
energy release in the prechamber causes the pressure there to rise above the main chamber
pressure . Depending on combustion chamber design and operating condition, the prechamber
pressure rises to be 0.5 to 5 atm above that in the main chamber. This pressure difference causes
a flow of fuel, air, and burning and burned gases into the main chamber, where the additional
energy release now occurs. The analysis of DI diesel in the previous section was base on uniform
pressure throughout the combustion chamber. For IDI engines the effect of the pressure
difference between the chambers must usually be included.

An IDI combustion chamber divided at nozzle into two open systems. Applying the first
law to the main chamber yields

dQ1 dV dm dU 1
p1 1 h2,1 (1)
dt dt dt dt

And to auxiliary chamber yields

dQ2 dm f dm dU 21
hf h2,1 (2)
dt dt dt dt

Here dm/dt is the mass flow rate between chambers with positive flow from the prechamber to
the main chamber. If dm dt > 0, h2,1 = h2 ; if dm dt < 0, h2,1 = h1 ;.If we define U 1 and U 2 as

dQ1 dQ2
sensible internal energies and h f as the sensible enthalpy of the fuel, then and
dt dt
represents the net energy release rate i.e. the difference between the combustion energy release
rates and the rates of heat transfer to the walls.

1
If we use an ideal gas model for working fluid in each chamber, with c p , c v and M

constant, the relation p1V1 m1 RT1 and p2V2 m2 RT2 can be used to eliminate m and T from

the dU dt terms and with the fact that hs , f = 0, can be used to write Eqs(1)and (2) as

dQ1 dV 1 dp dm
p1 1 V1 1 c pT2,1 (3)
dt 1 dt 1 dt dt

dQ2 1 dp dm
V2 2 c pT2,1 (4)
dt 1 dt dt

When Eqs(1)and (2) are added together, the term representing the enthalpy flux between the two
chambers cancels out, and the following equation for total net heat release results:

dQ dQ1 dQ2 dV 1 dp dp
p1 1 (V1 1 V2 2 ) (5)
dt dt dt 1 dt 1 dt dt

The comments made in the previous section regarding the interpretation of the net heat release

( it is the gross heat release due to combustion less the heat transfer to the walls, and other
smaller energy transfer due to crevices, fuel vaporization, and heat up) also hold here.

In practice , Eq(5) is difficult to use since it requires experimental data for both the main
and auxiliary chamber pressure throughout the combustion process. Access for two pressure
transducers through the cylinder head is not often available; even when access can be achieved,
the task obtaining pressure data from two different transducers under the demanding thermal
loading conditions found in IDI diesels, of sufficient accuracy such that the difference between
the pressures (of order 0.5 to 5 atm) at pressure levels of 60 to 80 atm can be interpreted, requires
extreme diligence in technique. Figure(1 and 2) shows apparent net heat release rate profiles for
IDI diesel obtained using Eq. (5) with = 1.35. Curves of dQ dt and dQ d are shown at three

2
different speeds and essentially constant fuel mass injected per cycle. While absolute heat
release rates increases with increasing speed, the relative rates are essentially independent of
speed, including that combustion rates, which depend on fuel-air mixing rates, scale
approximately with engine speed.

(1) (2)

Calculated net heat release rate profile for IDI diesel engine at constant load(0.29< <0.32)(1)heat release rate in kilowatts

(2)heat release rate in joules per degree

Eq(5) has been used assuming p 2 p1 and using either main chamber or auxiliary chamber
pressure data alone. The error associated with this approximation can be estimated as follows. If
we write p1 p2 p then Eq.(5) becomes

dQ dV V V2 dp1 V d (p)
p1 1 1 2 (6)
dt 1 dt 1 dt 1 dt

If the last term is omitted, Eq. (5) is identical to an Eq. derived for DI diesel. Since the term
V (dp1 dt ) ( 1) is much larger than the first term on the right hand side of Eq.(6) during early
stages of combustion process, the error involved in omitting the last term is given to a good

3
approximation by [V2 (V1 V2 )] d (p) dp1 In the initial stages of combustion this error can be
quite large . Later in the combustion process it becomes negligible. Thus , neglecting p will
lead to errors in predicting the initial heat release rate. The magnitude of the error will depend on
the design of the combustion chamber and on engine speed and load. Later in the combustion
process the error is much less, so integrated heat release data derived ignoring delta P will show
a smaller error.

A model analogous to the above, but using the approach of Krieger and Bormann, for the IDI
diesel has been developed and used by Watson and Kamel. The energy conservation equation for
an open system developed, with energy and enthalpy modeled using a consistent datum, with
appropriate models for convective and radiation heat transfer and for gas properties , was applied
to main chamber and also to the prechamber. These equations were solved using accurately
measured main chamber and prechamber pressure data to determine the apparent rate of heat
release( the rate of fuel burning multiplied by the fuel heating valve) in the main chamber and
prechamber through the combustion process. The engine was a Ricardo Comet swirl chamber

4
Calculated gross heat release rates in IDI swirl chamber diesel engine at full load (1) prechamber heat
release (2)main chamber heat release. Top fig: integrated heat release. Bottom figures: heat release rate

IDI design. Some results are shown in Fig. For this particular engine, at low engine speeds two-
third of heat release in the prechamber; at higher engine speed about two-third of the heat release
occurs in the main chamber.

5
Combustion in Direct Injection, Multi spray systems:-

Figure 10-6 shows typical data for cylinder pressure (p), fuel-injector needle-lift, and fuel
pressure in the nozzle gallery through the compression and expansion strokes of a direct-
injection diesel. There is a delay of 9 degreebetween the start of injection and start of combustion
[identified by the change in slope of the p()curve]. The pressure rises rapidly for a few crank
angle degrees, then more slowly to a peak value about 5" after TC. Injection continues after the
start of combustion. A rate-of heat-release diagramt from the same study, corresponding to this
rate of fuel injection and cylinder pressure data, is shown in Fig. 10-7. The general shape of the
rate-of-heat-release curve is typical of this type of DI engine over its load and speed range. The
heat-release-rate diagram shows negligible heat release until toward the end of compression
when a slight loss of heat during the delay period (which is due to heat transfer to the walls and
to fuel vaporization and heating) is

evident. During the combustion process the burning proceeds in three distinguishable stages. In
the first stage, the rate of burning is generally very high and lasts for only a few crank angle
degrees. It corresponds to the period of rapid cylinder pressure rise. The second stage
corresponds to a period of gradually decreasing heat-release rate (though it initially may rise to a
second, lower, peak as in Fig. 10-7). This is the main heat-release period and lasts about

1
40 .Normally about 80 percent of the total fuel energy is released in the first two periods. The
third stage corresponds to the "tail" of the heat-release diagram in which asmall but
distinguishable rate of heat release persists throughout much of the expansion stroke. The heat
release during this period usually amounts to about 20 percent of the total fuel energy. From
studies of rate-of-injection and heat-release diagrams such as those in Fig. 10-7, over a range of
engine loads, speeds, and injection timings, Lyn6 developed the following summary
observations. First, the total burning period is muchlonger than the injection period. Second, the
absolute burning rate increases proportionally with increasing engine speed; thus on a crank
angle basis, the burning interval remains essentially constant. Third, the magnitude of the initial
peak of the burning-rate diagram depends on the ignition delay period, being higher for longer
delays. These considerations, coupled with engine combustion photographic studies, lead to the
following model for diesel combustion. Figure 10-8 shows schematically the rate-of-injection
and rate-of-burning diagrams, where the injected fuel as it enters the combustion chamber has
been divided into a number of elements. The first element which enters mixes with air and
becomes "ready for burning" (i.e., mixes to within combustible limits), as shown conceptually by
the lowest triangle along the abscissa in the rate-of burning figure. While some of this fuel mixes
rapidly with air, part of it will mix much more slowly. The second and subsequent elements will
mix with air in a similar manner, and the total "ready-for-burning" diagram, enclosed by the
dashed line, is obtained. The total area of this diagram is equal to that of the rate-of-injection
diagram. Ignition does not occur until after the delay period is over, however. At the ignition
point, some of the fuel already injected has mixed with enough air to be within the combustible
limits. That "premixed" fuel-air

2
mixture (the shaded region in Fig. 10-8) is then added to the mixture which becomes ready for
burning after the end of the delay period, producing the high initial rate of burning as shown.
Such a heat-release profile is generally observed with this type of naturally aspirated DI diesel
engine. Photographs (such as those in Fig. 10-4a and b) show that up to the heat-release-rate
peak, flame regions of low green luminosity are apparent because the burning is predominantly
of the premixed part of the spray. After the peak, as the amount of premixed mixture available
for burning decreases and the amount of fresh mixture mixed to be "ready for burning" increases,
the spray burns essentially as a turbulent diffusion flame with high yellow-white or orange
luminosity due to the presence of carbon particles.

To summarize, the following stages of the overall compression-ignition diesel combustion


process can be defined. They are identified on the typical heatrelease- rate diagram for a DI
engine in Fig. 10-9.

3
Ignition delay (ab). The period between the start of fuel injection into the combustion chamber
and the start of combustion [determined from the change in slope on the p-8 diagram, or from a
heat-release analysis of the p() data.

Premixed or rapid combustion phase (bc). In this phase, combustion of the fuel which has
mixed with air to within the flammability limits during the ignition delay period occurs rapidly in
a few crank angle degrees. When this burning mixture is added to the fuel which becomes ready
for burning and bums during this phase, the high heat-release rates characteristic of this phase
result.

4
Mixing controlled combustion phase(cd).once the fuel and air which premixed during the
ignition delay have been consumed, the burning rate (or heat release rate)is controlled by rate at
which mixture becomes available for burning. While several processes are involved ;liquid fuel
atomization, vaporization, mixing of fuel vapour with air, preflame chemical reaction-rate of
burning is controlled in this phase primarily by the fuel vapor air mixing process.The heat release
rate may or may not reach a second(usually lower) peak in this phase; it decreases as this phase
progresses.
Late combustion phase(de).Heat release continues at a lower rate well into expansion stroke.
There are several reasons for this. A small fraction of the fuel may not yet have burned.A
fraction of the fuel energy is present in soot and fuel rich combustion products and can still be
released. The cylinder charge is nonuniform and mixing during this period promotes more
complete combustion and less dissociated product gases. The kinetics of final burnout processes
become slower as the temperature of the cylinder gases fall during expansion.

5
ANALYSIS OF CYLINDER PRESSURE DATA :-

Cylinder pressure versus crank angle data over the compression and expansion strokes of the
engine operating cycle can be used to obtain quantitative information on the progress of
combustion. Suitable methods of analysis which yield the rate of release of the fuel's chemical
energy (often called heat release), or rate of fuel burningis be described. The methods of
analysis are similar to thosefor spark- ignition engines and start with the first law of
thermodynamics for an open system which is quasi static (i.e., uniform in pressure and
temperature). The first law for such a system (see Fig. 9-11) is ,

6
dU
= + mihi -(1)
dt

where dQ/dt is the heat-transfer rate across the system boundary into the system, p(dV/dt) is the
rate of work transfer done by the system due to system boundary displacement, mi, is the mass
flow rate into the system across the system boundary at location i (flow out of the system would
be negative), hi is the enthalpy of flux I entering or leaving the system, and U is, the energy of
the material contained inside the system boundary. The following problems make the application
of this equation to diesel combustion difficult:
1) Fuel is injected into the cylinder. Liquid fuel is added to the cylinder which vaporizes and
mixes with air to produce a fuel air ratio distribution which is common unifom and varies
with time. The process is not quasi static.
2) The composition of the burned gases (also non uniform) is not known.
3) The accuracy of available correlations for predicting heat transfer in diesel engines is not
well defined.
4) Crevice regions (such as the volumes between the piston, rings, and cylinder wall)
constitute a few percent of the clearance volume. The gas in these regions is cooled to
close to the wall temperature, increasing its density and, therefore, relative importance of
these crevices. Thus crevices increase heat transfer and contain a non negligible fraction
of the cylinder charge at conditions that are different from the rest of the combustion
chamber.
Due to difficulties in dealing with these problems, both sophisticated methods of analysis and
more simple methods give only approximate answers.

Combustion Efficiency :-
In both heat-release and fuel mass burned estimations, an important factor is the completeness of
combustion. Thus, the combustion inefficiency is usually less than 2 percent, in other words,the
combustion efficiency is usually greater than about 98 percent .therefore it is a good
approximation to regard combustion and heat release as essentially complete.

7
Direct-Injection Engines :-
For DI engine, the cylinder contents are a single open system. The only mass flows across the
system boundary (while the intake and exhaust valves are closed) are the fuel and the crevice
flow.' An approach which incorporates the crevice flow has been described in Sec. 9.2.2; crevice
flow effects will be omitted here. Equation (10.1) therefore becomes ,

dU
= + mfhf (2)
dt

Two common methods are used to obtain combustion information from pressure data using Eq.
(2). In both approaches, the cylinder contents are assumed to be at a uniform temperature at each
instant in time during the combustion Process. One method yields fuel energy- or heat-release
rate; the other method yields a fuel mass burning rate. The term apparent is often used to
describe these quantities since both are approximations to the real quantities which cannot be
determined exactly.

HEAT RELEASE ANALYSIS


If U anfhfin(10.2) are taken to be the sensible internal energy of the cylinder contents and the
sensible enthalpy of the injected fuel, respectively, then dQ/dt becomes the difference between
the chemical energy or heat released by combustion of the fuel (a positive quantity) and the heat
transfer from the system (in engines, the heat transfer is from the system and by thermodynamic
convention is a negative quantity).since hs.f=0,eq..2 becomes,

dQn dQht dV
= = + /
dt

The apparent net heat-release rate, dQ/dt, which is the difference between the apparent gross
heat-release rate dQch/dt and the heat-transfer rate to the walls dQch/dt, equals the rate at which
work is done on the piston plus the rate of change of sensible internal energy of the cylinder
contents. If we further assume that the contents of the cylinder can be modeled as an ideal gas,
then Eq. (10.3) becomes

8
dQn dT dV
= +
dt

From the ideal gas law,PV=mRT, with R assumed constant, it follows that,

dV dP
. = . + .

dV dT
+ = . .

dP dV
= + . + .

dQn dP dV
= + . + .
dt

dQn + .
= + .
dt

dQn
= + . + .
dt


dQn/dt = / + /

1) is the ratio of specific heats,


2) for diesel heat-release analysis is 1.3 to 1.3
3) we would expect Y for diesel engine heat-release analysis to have values appropriate
to air at end-of-compression- stroke temperatures prior to combustion (= 1.35) and
to burned gases at the overall equivalence ratio following combustion (= 1.26-1.3).

9
More complete methods of heat-release analysis based on Eq. (10.2) have been proposed and
used. These use more sophisticated models for the gas Properties before, during, and after
combustion, and for heat transfer and crevic effects.8 However, it is also necessary to deal with
the additional issues of: (1 mixture nonuniformity (fuel air ratio non uniformity and burned and
unburned gas nonunifoties); (2) accuracy of any heat-transfer model used .

and (3) the effects of the crevice regions.


Additional insight can be obtained by incorporating a model for the largest of the effects omitted
from Eq. (10.6), the heat transfer dQht/ dt (we thereby obtain a close approximation to the gross
heat-release rate. The integral of the gross heat-release rate over the complete combustion
process should thenequal (to within a few percent only, since the analysis is not exact) the mass
of fuel injected m, times the fuel lower heating value QLHv i. e.,


Qch =
= mf. Q LHV (4)

Of course, Eqs. (1) to (2), (3), and (4) also hold with crank angleas the independent variable
instead of time t.

10
Figure 10-11 illustrates the relative magnitude of gross and net heat release, heat transfer, crevice
effects, and heat of vaporization andheating up of the fuel for a turbocharged DI diesel engine
operating in the mid-load, mid-speed range The net heat release is the gross heat release due to
combustion, less the heat transfer to the walls, crevice effects, and the effect of fuel vaporization
and heatup (which was omitted above by neglecting the mass addition term in dU/dt). This last
term is sufficiently small to be neglected. The enthalpy of vaporization of diesel fuel is less than
1 percent of its heating value; the energy change associated with heating up fuel vapor from
injection temperature to typical compression air temperatures is about 3 percent of the fuel
heating value. The heat transfer integrated over the duration of the combustion period is 10 to 25
percent of the total heat released.

FUEL MASS BURNING RATE ANALYSIS:


If the internal energies of the fuel, air, and burned gases in Eq. (10.1) are evaluated relative to a
consistent datum then this equation can be used to obtain an apparent fuel mass burning rate
from cylinder pressure versus crank angle data. such a species energy datum the "heat release" is
properly counted for in the internal energy and enthalpy terms.) Following Krieger and Borman
eq.10.2.can be written as,

d mu dQ dm
= + +
dt

Here Q is the heat transfer to the gas within the combustion chamber (that is, Q = -Qht,), m is the
mass within the ombustion chamber, and dm/dthas been substituted for mf. Since the properties
of the gases in the cylinder during combustion (assumed to be uniform and in chemical
equilibrium at the pressure p and average temperature T) are in general a function of p, T, and
the equivalence ratio

u=u(T,p,) and R=R(T,p,)


Therefore,
du d
= . / + / + /
dt

11
dR dR
= . + . + /
dt

+ /
= +
/


+


.

d/dt =

(F/A) is the fuel air ratio;the subscript 0 denotes the initial value prior to fuel injection and the
denotes the stoichiometric value.It then follows that ,


+

/ =
+
[ + / / ]

Where,

B = ( )( ) ( )( )( )+( )

/
C=
+ / /


[ + ]

D=

.

12
Equation (10.12) can be solved numerically for m(t) given m0,, 0 , p(t), and appro$ate models
for the working fluid properties (see Sec. 4.7) and for the heat $ate models for the working fluid
properties (see Sec. 4.7) and for the heat transfer term dQ/dt.
Figure 10-12 shows cylinder pressure data for an open chamber DI diesel and fuel mass burning
rate dm/dcalculated from that data using the above method. The heat-transfer model of Annand
was used (see Sec. 12.4.3). The result obtained is an apparent fuel mass burning rate. It is best
interpreted, after multi plying by the heating value of the fuel, as the fuel chemical-energy or
heat- release rate. The actual fuel burning rate is unknown because not all the fuel "burns" with
sufficient air available locally to produce products of complete combustion. About 60 percent of
the fuel has bumed in the first one-third of the total combustion period. The integral of the fuel
mass burning rate over the combustion process should equal the total fuel mass burned; in this
case it is 3 percent less than the total fuel mass injected. Note that chemical energy con- tinues to
be released well into the expansion process. The accuracy of this type o calculation then
decreases, however, since errors in estimating heat transfer sig significantly affect the apparent
fuel burning rate. Krieger and Borman also camed out sensitivity analyses for the critical
assumptions and variables. They found that the effect of dissociation of the product gases was
negligible. This permits a substantial simplification of Eq. (10.12). With no dissociation, u =
u(T, 4), and R = R/M can be taken as constant constant, since the molecular weight M changes
little. Then
13

dm [ + + /
=
dt +

/

WhereD, as before, is C 1 + (F/A)olm/[(F/A)m, ,]. Given the uncertainties inherent in the heat-
transfer model and the neglect of nonuniformities and crevices, Eq. (1) represents an adequate
level of sophistication. The other sensitivity variations studied by Krieger and Borman were:
shift- ing of the phasing of the pressure data 2" forward and 2" backward; translating the
pressure data f 34 kPa(5 lb/in2); changing the heat transfer +50 percent;

increasing the initial mass 5 percent.The mass change had a negligible effect on the fuel burning
rate calculations. The heat transfer changes of +-50 percent changed the mass of fuel burned by
about +-5%.The change in phasing of pressure data was more significant.It needs to be stressed
that accurate(in magnitude and phasing )pressure data are a most important requirement for
useful heat release or fuel burning rate analysis.
References:
1.Internal Combustion Engine Fundamentals By John B.Heywood.(page no. 503)
2.Internal Combustion Engines Applied Thermosciences By colin R Ferguson &Allan T.
Kirkpatrick
3. Gnter P. Merker Christian Schwarz Gunnar Stiesch Frank Otto Simulating Combustion
14
1

Analysis of a Single-Cylinder Diesel Engine for Non-linear Control

Theory:

Background :

The modelling of diesel engines has received much research attention. The models have
been used for engine simulation and design. In addition, with the recent rapid advance of
microprocessor technology, in the as become feasible to use these models for real time
applications, such as state estimation, control and fault diagnostics. However, there needs to be
further research in modelling techniques, to improve predictive accuracy, before models can be
used for estimation and control tasks. Several investigate or shave studied diesel engine the
thermodynamic modelling. The investigations can be divided for two categories:

i) Linear engine models relying on empirical data. These models link steady-state
experimental data representing engine thermodynamics and gas how, with simple dynamic
models of the mechanical components. The major disadvantage of such quasi-linear models is
the in heavy reliance on experimental data. Furthermore, the present of the transient response of
the engine.

ii) Non-linear engine simulation models using the Filling and Emptying Method, and the
Method of Characteristics, The set Thermodynamic engine models have been developed for
Engine design and performance predictions, and, few have been developed for diesel engine
control In a mean value model was used. The model consists of several empirical algebraic and
highest order differential equations in order to achieve real-time implementation. The main
drawback of this type of model is the requirement of empirical test data to The model equations.
Diesel engine models for electronic control development were developed in. The in-cylinder
combustion process was included by using a single zone model, and filling and emptying
2

modelling technique. In several medications to the model are proposed to cut down the
computational time and commercial rate cost. These models provide a tool for engine simulation
but is not suitable for diagnostic formulations and in-cycle crank shaft angular speed variations
prediction.

A non-linear model for a single cylinder diesel engine is developed. The model can be
used as an engine simulator to aid diesel engines control system design and fault diagnostics. The
model does not use any empirical inputs, and accurately predicts in cycle variations of engine
parameters during transient response.
The model treats the cylinder strokes and the manifolds as thermodynamic control
volumes by using the modelling and emptying method, solving energy and mass conservation
equations with sub-models for combustion, heat transfer, instantaneous friction and dynamic
analysis. Inertia variations and piston pin offset are taken in to consideration. In cycle
calculations are performed at each crank angle, so the correct crank-angle of ignition, speed
variation, amount of fuel and air as well as the fuel burning rate are predicted. The model is
validated using experimentally measured cylinder pressure and engine instantaneous speeds,
under transient operating conditions, giving good agreement.

Model Analysis:-
A general engine model will consist of both thermodynamic and dynamic equation. The
model inputs will be the air and fuel flow rates, and the load on the engine. Then, given the
defined engine parameters, the model should predict the crankshaft speed variations. The
thermodynamic model presented in this paper is based on the filling and emptying method. Three
thermodynamic control volumes are considered; the inlet manifold, the cylinder and the exhaust
manifold. A schematic is shown in Figure. Equations for the conservation of mass and energy,
gas property relations and perfect gas law are solved in the crank-angle domain. A single zone
combustion model is used for simulating the theoretical fuel burning rate. The cylinder is treated
as one control volume with homogeneous temperature and pressure. The non-linear dynamic
model in, including instantaneous friction components model is improved to include the
computation of the inertia variation of crankshaft assembly, and the time domain is transformed
to the crank angle domain.,
3

Equations for Conservation of Mass and Energy :

The filling and emptying method treats the cylinder and each manifold as a
thermodynamic control volume. The first law of thermodynamic for an open system can be
written as:

where subscript i denotes different entries to the control volume and () represents

derivatives with respect to crank angle . Above Equation can be written as:

where j denotes surfaces with different rates of heat transfer, Q includes heat released by
combustion, ho is the stagnation enthalpy, hence kinetic energy implicitly accounting for, u is the
specific internal energy, P and T are control volume pressure and temperature respectively and m
is the mass.

Differentiating the left-hand side of above Equation gives

The specific internal energy u is assumed to be a function of temperature T and the


equivalence ratio F only. Since dissociation effects are small, the influence of pressure is
neglected. Thus,

u=u(T,F)
A curve fit of combustion product data, as a continuous function is used [4], based on
absolute zero temperature of the elements:

Where,
4

And

The gas constant R is given by,

where fs is the stoichiometric fuel air ratio.


This technique has the advantage of automatically predicts a reduction of specific heat
release by combustion at richer than stoichiometric fuel air ratio, which occur during transient
operation. Thus,

where Q0 ht is the heat transfer rate, mf is the mass of fuel burnt, and h for is the enthalpy
of formation of the hydrocarbons (fuel), given by :

hfor = 2:326[ 19183+ .5(Tfuel 537) ]

is the low heating value of the fuel. Assuming that the gases involved behave as perfect
gases, PV = mRT, where R is the gas constant. Using these values above Equations becomes

Mass conservation for both air and fuel is required. Considering the total mass (fuel and
air) gives,

The total mass is given by,

where ma is the mass of the air and mbf is the mass of the burnt fuel. The fuel-air
equivalence ratio is defined as,
5

where f is the fuel air ratio. Thus, the mass of the burnt fuel mbf in total mass m of air
and burnt fuel is given by

From Equations, the term F in above Equation can be found;

Equations for Ports and Valves Mass Flow Rates:

A one-dimensional model for flow through a valve (or port) using the analogy of an
orifice having and equivalent area, is used. Applying the energy equation from upstream to
downstream for isentropic steady flow, and assuming that the inlet velocity is negligible, for

subsonic flow
The mass flow rate has the form as below

Where Cd is the discharge coefficient, is the specific heat ratio and Av is the valve or port area.

For sonic flow the mass flow rate has the form,
6

Equations for Combustion:

Many combustion models have been discussed in. The single zone model proposed in is
used here, because it is widely accepted and can be used to predict both premixed and diffusion
burning. The ignition delay (ID) equations as function of cylinder pressure and temperature can
be used to calculate the start of the combustion.

where inj is the angle at injection and ign is the angle at ignition The normalised premixing
burning rate is given by:

while the normalised diffusion burning rate is given by:

where norm is non-dimensional angle, increasing from 0 at start of combustion to 1 at


the end of combustion. The parameters kp1, kp2, kd1 and kd2 have been correlated with
fundamental factors having the strongest influence on the combustion rate. The best fit values of
these parameters for a range of direct injection diesel engines was presented in,
7

Equations for Cylinder Heat Transfer :

An instantaneous convective heat transfer is defined by,

where hcon is the heat transfer coefficient based on forced convection and is given by .

Np is the mean piston speed and Twall is the cylinder wall temperature. The primary
sources of radiative heat transfer in a diesel engine are the high temperature burned gases and
soot particles. Following work in, the instantaneous radiant heat transfer can be expressed as:

where is the apparent grey-body emissivity, is Stephan-Boltzman constant and Trad


is the apparent radiant temperature. The adiabatic flame temperature of the slightly greater than
stoichiometric (F = 1:1) zones of hydrocarbon-air combustion products. The apparent radiant
temperature is calculated as the mean of the adiabatic flame temperature and the average bulk
gas temperature, i.e.
8

The temperature of combustion products at F = 1:1 is computed from a correlation of


instantaneous air temperature and pressure as follows

for 800K < T < 1200K, and

for 450K < T < 800K

The apparent emissivity is assumed to vary linearly between its maximum value :9 to
zero over the expansion stroke. The cylinder wall temperature is updated at each step by using
the one-dimensional heat conduction model, and an electrical analogy model, shown in Figure :

Where Rwc = 1 hcoolant A is the thermal resistance from wall to coolant. The thermal
resistance for the conduction through the wall is Rw = t KA, where t is the wall thickness and K
is the thermal conductivity. Then the fourth order polynomial equation in Tm as an independent
variable is obtained and the solution gives the wall temperature.
9

Equations for Exhaust Manifold Heat Transfer:

The exhaust gas flow is considered turbulent. The forced heat convection between the
exhaust gas and the manifold wall, the natural convection, and radiation between the manifold
and ambient air are taken into account. The conductive transfer through the wall thickness is
neglected. An electrical analogy model is shown in Figure 3. The manifold is assumed to have a
cylindrical shape, homogeneous gas and wall temperatures. The heat transfer rate from exhaust
gas to manifold is given by,
Qht = hfAm(T Tm)

The heat transfer rate from manifold to air is given by;

Qma = hnAm(Tm Tair)

The heat transfer rate due to radiation is given by,

Qr = Am(T4m T4air)

where hf is the forced heat convective coefficient given by

where Re is the Reynolds number, Pr is the Prandtl number:


10

is the kinematic viscosity, Di is the inner manifold diameter and Cp is the specific heat.

The dynamic viscosity can be calculated from :

The natural heat convective coefficient hn is given by,

where Gr is the Grashof number

where g is the gravitational acceleration, Do is the outer manifold diameter and & is the
volumetric coefficient and it is taken equal 1 Tair . Applying the energy conservation principle to
the volume, a non-linear algebraic equation in Tm is obtained.
The solution to this equation gives the manifold temperature which can be substituted
into Equation 38 to calculate the manifold heat transfer rate.

Cylinder, Intake and Exhaust Manifolds Modelling:

The solutions of above Equations are considered for cylinder, inlet and exhaust manifolds
in turn. In all cases it is assumed that T; m; F, and P are known at the beginning of each step. For
the first step, estimate values are used; for the remaining the values steps are known from the
previous step.

Cylinder:
11

i) Compression stroke, from close of the intake valve to open of the exhaust valve:
During this period F = 0; m = 0; h inorout = 0; mfb = 0 Equation becomes

ii) Combustion and expansion: During this period m = m f ; hinorout = 0


Equations become

iii) Exhaust stroke: During this period, combustion is assumed to be completed and
the equivalence ratio is constant. i.e F = 0, m in = 0, hin = 0, mf = 0 and m =
m0 out: Equations becomes

iv) Intake stroke: During this period, mout = 0, mf = 0, m fb = 0 and m= min:


Equations becomes

Inlet Manifold:
12

The air flow is supplied, hence the equivalence ratio is zero, so F= 0. In addition m fb =
0, Qht = 0, V = 0 and m = m in mout: Equation becomes

Exhaust manifold :

Here, m fb = 0, V = 0 and m = m inm out: above Equations becomes

Model implementation:
The model is implemented using MATLAB/SIMULINK. The main advantage of
SIMULINK is its capability for representing the entire engine model by an assemblage of
interconnected blocks. Also it has eight variable-step solvers and six fixed-step solvers for the
integration of differential equations, and hence the most suitable integration method can be
chosen. Input design parameters are passed on to the blocks from an input _le, but all of the
operating parameters come from the block (functions) for the other components of the system.
Fourth order Runge-Kutta numerical integration is used to solve the differential equations for the
detailed model. The integration step for the program is chosen to be one crank-angle degree. The
indicated torque from the cylinder is updated at each step, and then used in the non-linear
dynamic model to update the instantaneous crankshaft angular velocity. The current simulation
program is executed on a (400 MHZ) Pentium II PC, and the simulation takes about 8 seconds
for a 720 degree engine cycle at an engine speed of 950 rpm.
13

Model behaviour and validation:


In order to validate the behaviour of the engine model with experimental results,
simulations were per- formed for a F1L 210 DEUTZ MAG DI single cylinder diesel engineA
comparison between predicted and measured values of indicated pressure and crankshaft
instanta- neous angular velocity during engine transient behaviour is illustrated in Figure 5 and
Figure 6 respec- tively. The measured values are taken from . Figure shows a comparison
between the predicted and measured cylinder gas pressure with _ring at an average engine speed
of 920 rpm and 25 mm3 of fuel per cycle. The overall agreement between the measured and
predicted curves is good. Owing to the measured data of the gas pressure from the engine
without an exhaust manifold, the gas pressure during the exhaust stroke is overestimated. The
slight underestimation during the combustion stroke may be due to the fact that a single zone
combustion model is used here; a slight difference in engine timing between the practical and
simulated one and a minor change between the real heat transfer, heat release and the simulated
one.

The slight underestimation of the gas pressure during combustion has resulted in an
underestimation of the instantaneous engine speed near TDC. Also the underestimation of
instantaneous engine speed during the exhaust stroke is due to the gas pressure overestimation as
shown in Figure 6. In addition, there are another reasons like, inaccuracies in the values of the
engine model parameters and changes in friction behaviour. The overall agreement between the
measured and predicted curves is good. The normalized rate of heat release is shown in Figure 7.
The high left peak in the curve represents the premixed phase. Combustion of fuel which has
mixed with air to within the amiability limits during the ignition delay period occurs rapidly in a
few crank angle degrees. When this burning mixture is added to the fuel it burns during this
phase, and a high heat release rate is characteristic of this phase. The low right peak is the
diffuusion phase; once the fuel and air premixed during the ignition delay have been consumed,
the burning rate is controlled by the mixture available for burning. The heat released decreases as
this phase progresses as shown in Figure 7. The instantaneous torque produced by the engine at
40 mm3 of fuel per cycle for four engine cycles is shown in Figure 8. The maximum torque
value represents the maximum pressure in the cylinder during the combustion stroke as
14

illustrated in the same figure. As a consequence of the fluctuations in engine torque during the
cycle, variations of the instantaneous crankshaft rotational speed occur.
Conclusions:

A dynamic dynamometer model is also included, which enables a variety of engine tests
to be carried- out. In addition, the model is based in the crankshaft angle domain. This type of
model is a better candidate for model-based diagnostics since oscillatory engine states are
intrinsic in the model and can be related to engine failures. Also, inclusion of the inertia variation
effects, especially at high engine speeds; significantly decrease the occurrence of false diagnostic
alarms due to more accurate pressure estimation. Future advances in microprocessor speed will
allow the use of such those non-linear models for estimation and control in real time. The model
has been developed with the aim of investigating different strategies for transient fuel control.
The model is being extended for multi-cylinder engines. Work is also ongoing to include
modelling of engine operation from cold start.

References

1. Complete Analytical Model of a Single-Cylinder Diesel Engine for Non-linear Control and
Estimation by Yahya H. Zweiri.
2. John B Heywood, Internal combustion engine fundamental.
Compressor Map Digitization and Performance Prediction Using Matlab and
Neural Network Tool in Matlab

Abstract

In this paper, matlab programs are made to predict performance of the compressor in terms of
efficiency using compressor map.A compressor map of Honeywell GT2860RS compressor is
digitized using digitizing tool and a digitized map is made inMatlab.Out of four parameters on
compressor map two are taken as inputs to the program and remaining two are
calculated.Different combinations of inputsand outputs are termed as models. For fitting the
constant speed lines andconstant efficiency contours different methods like curve fitting, ellipse
fitting, artificial neural networks are considered and compared. The results indicate that the
methods used in the programs of various models have acceptable root mean square error.

Keywords: compressor map, curve fitting, contour fitting, neural network

1. Introduction

The compressor map is a graphical representation that describes a particular compressors


performance characteristics, including efficiency, mass flow range, boost pressure capability, and
turbocharger speed. Surge line is the left hand boundary of the compressor map. Operation to the
left of this line represents a region of flow instability andmild flutter to wildly fluctuating boost
from the compressor. Long operation within this region can lead to premature failure due to
heavy thrust loading. Choke line is the right hand boundary of the compressor map. Operation to
the right of this line causes rapid drop of compressor efficiency and the turbocharger speed will
also be approaching or exceeding the allowable limit. If operation is beyond this limit, a larger
compressor is necessary. Thus it is very essential to predict the performance of compressor for
operating range of speed, mass flow rate, pressure ratio etc. Based on the estimated data suitable
compressor can be selected for the application without conducting actual test on the compressor.
For the estimation or prediction of the performance parameters standard compressor maps can be
used.
This work is motivated by,An overview of curve fitting methodsfor centrifugal compressor
and turbine characteristics Moraal and Kolmanovsky [1] presented the curve fitting methods for
compressor and turbine maps.An artificial neural network approach to compressor performance
prediction Ghorbanian and Gholamrezaei [2] presented comparison across different ANNs in
predicting axial compressorperformance map.Improved compressor maps using approximate
solutions to theMoore Greitzer modelChris DrummondandCraig R. Davison [3] presented use of
expertise knowledge for smoothing of speed curves and efficiency contours. Compressor
Modeling for Control of Automotive Two Stage Turbochargers Oskar Leufven [4-6] presented
3D and 1D compressor models,physical 0D compressor models,curve fitting 0D based models.A
compact and accurate empirical model for turbine mass flow characteristicsXiande Fang, Qiumin
Dai, Yanxin Yin, Yu Xu [7] presented a mean value model for turbine mass flow rate prediction
with very low mean absolute relative error.

The aim of this work is to make matlab programs for predicting the performance by the use
of some digitized map of the physical map. Thedigitized map data can be used for interpolation
between the speed, efficiency curves and calculation of the output parameters. The curve fitting
is one of the approaches that recognize all performance variables in terms of constant speed lines
and the efficiency contours of the map. The modeling effort is then to fitdifferent curves to the
digitized map.Different approaches to capture the characteristics of the compressor map are
presented.Using this approaches different input output models for a compressor map are made.

Model 1:, = , (1)

Model 2: , = , (2)

In this formulation, one may obtain one single network with two inputs and two outputs.
However, analysis reveals that both Eqn (1) and Eqn(2) is not accurate enough [1]. In this regard
a different formulation for the determination of the compressor efficiency as follows:

= , (3)

Thus,

Model 1 , = , is splitted into

Step 1: Pressure ratio is calculated from input mass flow rate and speed
= ,

Step 2: Efficiency is calculated from mass flow rate and pressure ratio.

= ,

Model 2 , = , is splitted into

Step 1: Mass flow rate is calculated from input pressure ratio and speed

= ,

Step 2: Efficiency is calculated from mass flow rate and pressure ratio.

= ,

Here we can see step 2 for both models are same that is efficiency is calculated as a function of
Ma and P. So this part is common in both models.So the complete models becometwo inputs two
output program.

Model 1: Inputsas mass flow rate and speed having outputs as pressure ratio and efficiency.

Model 2: Inputs as pressure ratio and speed having outputs as mass flow rate and efficiency.

A third model is also possible having inputs mass flow rate and pressure ratio and outputs are
speed and efficiency but as speed of compressor is not characteristic of compressor it is
characteristic of engine or turbine output hence model 3 is not of interest.

2. Compressor map transformation in to matlab

2.1 Digitization of map

Digitization is process of taking various coordinates of points from the given physical graph.
This can be done using various software tools available for digitization of graph. For our purpose
we have used Web Plot Digitizer which is open source online graph digitizing tool. The
coordinates mass flow rate along x-axis and pressure ratio along y-axis for various constant
speed curves, constant efficiency contours, surge line, choke line are obtained and saved as row
arrays. The data of mass flow rate, pressure ratio, speeds, and efficiencies are saved in proper
ways such that they are available for further operations such as curve fitting, interpolation etc.

2.2 Curve fitting using different degree of polynomial

Curves can be fitted to the various points obtained by digitization using polyfit and polyval
commands in matlab. In this commands a certain degree of polynomial is fitted to all the points.
Thus we can select the degree of polynomial for which fitting of curve is close to all the points
that is smoothing of the speed curve [3] The curve fitting using different degree polynomial is
shown in Fig. 1

Fig. 1 Curve fitting using different degree


polynomial

From the above observation we conclude the degree 4 and 5 gives best fitting to all the
points. Degree 5 requires more number of points hence we select degree 4 polynomial for our
purpose.

2.3 Transformed map

Using above method various curves are fitted for digitized points of compressor map and
transformed map is obtained. The actual map Fig. 2 (a) and digitized map Fig. 2 (b) is shown
below
Fig. 2 (a) Actual map (b) Digitized map

3. Methods used for different models.

3.1 Calculation of pressure ratio for given input mass flow rate and speed. = , :
The compressor map consists of constant speed lines for various mass flow rate and pressure
ratio. So firstly the speed is checked for two adjacent constant speed curves. Corresponding
speed line is plotted then pressure ratio for given mass flow rate is obtained by interpolation
between two constant speed lines. Hence pressure ratio for input mass flow rate and speed is
calculated. For example: Inputs Ma=24, N=130000, the calculated constant speed curve and
operating point is shown in Fig. 4 The output pressure ratio as 2.285492.
Fig. 4 Calculated speed curve , operating point Fig. 5 Error estimation in PR calculations

Pressure ratios are calculated and plotted by program for a set of 87 known values of mass
flow rate, pressure ratio on various costant speed curves and root mean square between estimated
values and actual kown values is calculated denoted by RMSE.The plot for estimated pressure
ratio and actual pressure ratio shown in Fig. 5 the estimated values are acceptable with *RMSE
of 0.0006909.

2
* =

where, n= number of sample values,

RMSE is used because the calculated values may lie on either side of the actual value i.e., error
may be positive or negative hence to take in account all error values summation of squares of all
errors is done, its mean is calculated by dividing it by number of samples then squareroot is
taken to get root mean square error (RSME). Here after RMSE is used to validate the prediction
of the program.

3.2 Calculation of mass flow rate for given input pressure ratio and speed. = , :

For input speed a constant speed line is determined and plotted then the corresponding mass
flow rate is obtained by interpolation on the calculated speed line. For example: Inputs Pr=1.8,
N=116000, the calculated constant speed curve and operating point is shown in Fig. 6 The
output mass flow rate as 28.860758.

Fig. 6 Calculated speed curve, operating point Fig. 7E rror estimation in Ma calculations

Mass flow rates are calculated and plotted by program for a set of 44 known values of mass
flow rate, pressure ratio on various costant speed curves and root mean square between estimated
values and actual kown values is calculated denoted by RMSE. The plot for estimated mass flow
rate and actual mass flow rate is shown in Fig. 7. The estimated values are acceptable with
RMSE of 0.053739

3.3 Calculation of efficiency = ,

The compressor map consists of constant efficiency contours. A line passes through picks of
each contour called as backbone line. These contours can be fitted in different ways. First way is
to fit the contours as curves shown in Fig. 8, in this method the contours are divided into two
parts, above backbone line and below backbone line. Separate curves are fitted for them using
polyfit command. The calculated mass flow rate and pressure ratio from first step is checked
conditionally between the saved values for each constant efficiency curves and two curves
between which this point lies are determined. Then interpolation is done between the two curves
to get efficiency of the operating point. Secondway is to approximate the contours as ellipses of
different sizes, in this method ellipses are fitted for set of points of the contours using matlab
programs available for ellipse fitting on internet in math works library. The programs fits ellipse
passing through the scattered points and gives length of axes, eccentricity, inclination of the axis
as output. By running the program for different contours the length of axes, inclination etc. is
obtained and imported in our program. To verify the fitting of ellipse they are plotted on the
transformed map as shown in Fig. 9Corresponding lengths of major axes of ellipse are used for
interpolation to get the efficiency of operating point.

Fig. 8 Curve fitting for contours Fig. 9 Ellipse fitting for contours

The methods above have some disadvantages in curve fitting the points where efficiency
contours are discontinuous are unable to take into account while in ellipse fitting the contours
does not have common center and eccentricity so the interpolation based on ellipse fitting are
very much approximate which can have very much error for some input cases. Hence both the
approaches are not taken for efficiency calculation.
Reliable and quick methods must be implemented for the efficiency calculation which can be
done using artificial neural network tool in matlab. Matlab has neural network tool (nntool) to
design train and simulate various neural networks according to problem definition. For our
purpose curve fitting networks are used. There are different methods to design the network.
Some of them are general regression neural network (GRNN), radial basis function network
(RBFN), design exact radial basis network (RBE) and Function fitting neural network (FFNN).
The capability of different types of networks to find efficiency from digitized data of compressor
map is investigated. The accuracy of the predicted values by various networks is compared with
actual values.Comparison of various neural networks is done in terms of root mean square error
(RSME). The result obtained is as follows:

Table 1: RMSE in efficiency calculation using different networks

Sr No. Network RMSE

1 Radial basis function network (rb) 0.0061721

2 design exact radial basis network (rbe) 0.0062593

3 general regression neural network (grnn) 0.15731

4 Function fitting neural network (fitnet) 0.02158

The table shows root mean square error for rb and rbe network is least among all the networks.
In addition to that rbe is faster than rb network hence rbe is selected for model of calculation of
efficiency.

Syntax for creating the network is

net = newrbe (P, T, spread)

Where, P and Tismatrix of input and target vectors can take 2 or 3 arguments, spread is spread of
radial basis function, the larger the spread is, smoother the function approximation will be. Too
large spread can cause numericalproblems default spread is 1.00. newrbecommand very quickly
designs a radial basis network,sim command is used for simulation of required input values with
in the network. Output of the network simulation is efficiency of compressor according to input
parameters.
For example: Inputs Ma= 30, Pr= 2.2, rbe network is formed and simulated for efficiency to
give output as 77.208233. The operating point on maps shown in Fig. 10
Fig. 10 Efficiency relating to operating point Fig. 11 Error estimation in c calculations

Efficiencies are calculated and plottedby program for a set of known values of mass flow rate
and pressure ratio on costant efficiency contours and root mean square error between estimated
values and actual known values is calculated.The plot for estimated efficiency and actual
efficiency is shown in Fig. 11. The estimated values are acceptable with RMSE of 0.0062593.
Set of continuous inputs according to varying turbine speed is given to the program the
output of program gives a curve passing through all operating points on compressor map which
is called as operating curve shown in Fig. 12 and Fig. 13
Fig. 12 Output curve for Model 1 Fig. 13 Output curve for Model 2

From the output of the program we can predict whether the compressor is able to handle all
operating points can determined, if not then other compressor is to be used.

4. Result and summary

Table 1 Result and summary of variousmethods used in models of compressor performance


prediction.

No. Formulation Methodsused RMSE Remark Application

Curve fitting for fitting


a smooth curves for
speed lines and
interpolationto get
Step 1 of model
1 = , intermediate speed 0.0006909 Acceptable
1
curve between fitted
constant speed curves.
Interpolation to get Pr
for input Ma and N.
Curve fittingfor fitting
a smooth curves for
speed lines and
interpolation to get
Step 1 of model
2 = , intermediate speed 0.053739 Acceptable
2
curve between fitted
constant speed curves.
Interpolation to get Ma
for input Pr and N.

Not sufficient
Curve fitting - -
for all inputs

Not sufficient
Ellipse fitting - -
for all inputs
3 Neural Networks
= ,
rb 0.0061721
rbe rbe network Step 2 of model
0.0062593
grnn acceptable 1 and model 2
fitnet 0.15731

0.02158

Curve fitting,
, Combining 1 and 3
interpolation within
4 = , Two input two output program for
fitted curves and
compressor map performance prediction
Neural network

Curve fitting,
Combining 2 and 3
, interpolation within
5 Two input two output program for
= , fitted curves and
compressor map performance prediction
Neural network
5.Conclusion

The matlabprogram is two inputs and two output function which give performance prediction
of compressor in terms of efficiency and position of operating point on the map. Curve
fitting,interpolationwithin fitted curves, and neural network are employed to calculate unknown
parameters at desired positions on compressor map.It should be emphasized that while the above
mentioned methods assist in reducing the time and cost of actual testing, they remain prone to
errors for performance prediction. The present work investigates the capability of curve fitting
and neural network tool inMatlab to reconstruct thecompressor map from digitized data. Results
show that estimation of pressure ratio using curve fitting results in RMSE of 0.06909%
estimation of mass flow rate using curve fitting results in RMSE of 0.053739% and estimation of
efficiency by design exact radial basis neural network results in RMSE of 0.66773%.

References

[1] Moraal P. and Kolmanovsky I., Turbocharger Modeling for Automotive Control
Applications, SAE Technical Paper 1999-01-0908, 1999, doi: 10.4271/1999-01-0908.

[2] K. Ghorbanian, M. Gholamrezaei, An artificial neural network approach to compressor


performance prediction, Science direct Applied Energy 86 (2009) 12101221.

[3] Chris Drummond and Craig R. Davison, Improved compressor maps using approximate
solutions to the Moore Greitzer model, Proceedings of ASME Turbo Expo 2009: Power for
Land, Sea and Air June 8-12, 2009, Orlando, Florida, USA.

[4] Oskar Leufvn, Compressor Modeling for Control of Automotive Two Stage
Turbochargers, Linkping Studies in Science and Technology Thesis No. 1463 Department of
Electrical Engineering Linkpings universitet, SE581 83 Linkping, Sweden

[5] Oskar Leufven and Lars Eriksson, Time to surge concept and surge control For acceleration
performance , Vehicular Systems, Department of Electrical Engineering, Linkping University,
S-581 83 Linkping, Sweden.

[6] Oskar Leufven and Lars Eriksson, Engine Test Bench Turbo Mapping , Vehicular Systems,
Department of Electrical Engineering, Linkping University, S-581 83 Linkping, Sweden.
1

ANALYSIS OF TURBOCHARGED AIR STANDARD DUAL CYCLE

ABSTRACT
The power output of an engine depends upon the amount of air inducted per unit time, the degree
of utilization of this air and thermal efficiency of the engine. The amount of air inducted can be
increased by increasing the engine speed or by increasing density of air at intake. The increase in
engine speed calls for rigid and robust engine as the inertia loads increase. The engine friction
and bearing load also increases and the volumetric efficiency decreases when the speed is
increased. The method of increasing the inlet air density is called Supercharging. It is mainly
employed to increase power output of engine.
The power output can also be increased by increasing the thermal efficiency of the engine by
increasing the compression ratio. However this increases the maximum cylinder pressure. The
rate of increase of maximum cylinder pressure is less than rate of increase of break mean
effective pressure in case of supercharged engine. This means that for a given maximum cylinder
pressure more power can be obtained by increasing compression ratio. The rate of increase of
maximum temperature is also low in case of supercharging.

INTRODUCTION
Supplying the pressure higher than the pressure at which the engine naturally aspirates air from
the atmosphere. The greater the density of an air-fuel charge forced into the cylinder, the greater
the force and engine power produced when ignited.
The increase in the amount of air inducted per unit time by supercharging is obtained mainly to
burn a greater amount of fuel in a given engine and thus increase its power output. The objects of
supercharging include one or more of the following:
1. Air density decreases with the altitude and hence the power output. This negative effect can be
controlled by using supercharger.
2. Volumetric efficiency can be controlled by supercharging.
3. Supercharging can be very effective to improve scavenging efficiency that controls the
performance of two-stroke engines.
2

TURBOCHARGING OF CI ENGINES:

Unlike SI engines supercharging does not result in any combustion problem rather it improves
combustion in a diesel engine. Increase in pressure and temperature of the intake air reduces
ignition delay and hence the rate of pressure rise resulting in a better, quieter and smooth
combustion. This improvement in combustion allows a poor quality fuel to be used in a diesel
engine and it is also not sensitive to the type of fuel used. The increase in intake air temperature
reduces volumetric and thermal efficiency but the increase in the density due to pressure
compensates for this intercooling is not necessary for highly supercharged engine.
However, mechanical and thermal loading increases with an increase in supercharging. But this
increase in mechanical and thermal loading is only moderate because of the lower compression
ratio and effect of cooling due to increased valve overlap of the supercharged engine. It is
possible to use the lower fuel-air ratio in the supercharged engine as the increase in fuel flow is
less than increase in air flow. This results in lower temperatures over the full engine cycle and
reduced smoke from the engine. Low fuel-air ratio and high expansion ratio result in lower
exhaust temperature. This result in increased life of exhaust valves.
If an unsupercharged engine is supercharged it will increase the reliability and durability of the
engine due to smoother combustion and lower exhaust temperatures. The degree of
supercharging is limited by the thermal and mechanical load on engine and strongly depends on
the type of supercharger used and design of the engine.

TURBOCHARGING LIMITS OF CI ENGINE :


Unlike SI engines, the limits of supercharging for compression ignition engines are not due to
combustion. The factors which limit the output of a SI engine due to knock result in quieter and
smoother operation of a CI engine. Any increase in pressure and temperature decreases the
ignition delay and engine runs smoother.
The limits of supercharging for a CI engine are reached by a thermal loading. The temperature of
the piston and cylinder if very high results in scuffing of piston rings and heavy liner wear. The
3

load on bearing is increased due to increased pressure in the cylinder. A maximum cylinder
pressure of about 85 bar limits the supercharger pressure to about 2 atmospheres with a
compression ratio 15:1 if a copper-lead bearing is used.
Durability, reliability and fuel economy are the main considerations in limiting the degree of
supercharging of a CI engine. Increase in maximum pressure in cylinder decreases the reliability
of the engine. This also increases the rate of heat release and hence the thermal load on the
engine is increased. For intake pressures greater than 2.5 atmosphere very sturdy and well cooled
engines are needed. Fuel economy is also an important consideration in deciding the degree of
well cooled engines is needed.
Fuel economy is also an important consideration in deciding the degree of supercharging for a CI
engine. For intake pressure less than 1.5 atmosphere the cost and complication of the
supercharger is not justified.

Now we will discuss performance of dual engine with turbocharging:

Vs+Vc
= Vc
(Volumetric compression ratio)

Pz
= (Pressure ratio)
Pc

Vz
= Vc
. (Cut-off ratio)

Vg
= Va (Coefficient of cooling in pre-cooler)

Va
c = Vk
(Volume ratio for compressor)
4

Cp
= .. (Adiabatic index)
Cv

Vk
= . (Coefficient of cooling in aftercooler)
Va

Expression for efficiency:

Work Done
=
Heat Supplied

Heat added Heat rejected


=
Heat added

Heat rejected
=1
Heat added
Cp Tg Ta + Cp Tk Ta
= 1
Cv Tz Tc + Cp Tz Tz

Let each temperature be expressed in terms of Ta

For Tg
PgVg PaVa
=
Tg Ta

But
Pa = Pg
Vg
Tg = Ta
Va
Tg = Ta (1)
Now

Ta Vk 1
=( )
Tk Va

1 1
Ta = Tk c
(2)

Now Tk
5

PkVk PaVa
( ) = ( )
Tk Ta

Vk
Tk = Ta
Va
Tk = Ta (3)

From (2)
1 1
Ta = Ta (4)
c
1 1
Tg = Ta c
(5)
PzVz PcVc
( ) = ( )
Tz Tc

Pz
Tz = ( ) Tc
Pc

= Tc (6)

For Tc;
Tc Va 1
( )=( )
Ta Vc

Tc = Ta 1 (7)
From (6)
Tz = Ta 1 (8)
For Tz;
PzVz PzVz
( )=( )
Tz Tz
Vz
Tz = ( ) Tz
Vz

= Tz
From (8)
Tz = Ta 1

1 1 1 1
Cp [( Ta c Ta c ) + Ta Ta ]
= 1
Cv Ta 1 Ta 1 + Cp Ta 1 Ta 1
6

[ 1 + c 1 1 ]
= 1
c 1 [ 1 + 1 ]

= f
Expression of Tg in terms of Ta;
Tg is expressed in terms of Ta by two routes,
We get;
Tz Vg 1
( )=( )
Tg Vz
Vz 1
Tg = ( ) Tz
Vg
PzVz PzVz
=( )
Tz Tz
Vz
Tz = ( ) Tz
Vz
PzVz PcVc
=( )
Tz Tc

Pz
Tz = ( ) Tc
Pc
Ta Vc 1
=( )
Tc Va
Vz 1 Vz Pz Va 1
Tg = Vg Vz Pc Vc
(1)

Tg in terms of Ta by another route

PgVg PaVa
=
Tg Ta
Vg
Tg = ( ) Ta
Va
Ta Vk 1
=( )
Tk Va
Vk 1
Ta = ( ) Tk
Va
PkVk PaVa
=
Tk Ta
Vk
Tk = ( ) Ta
Va
Vg Vk 1 Vk
Tg = Ta (2)
Va Va Va
7

Equating equations (1) and (2)

Vz 1 Vz Pz Va 1 Vg Vk 1
( ) ( )( )( ) = ( ) ( ) ( )
Vg Vz Pc Vc Va Va
1 1 1
Vz Va Vk
( ) ( ) = ( )
Vg Vc Va
Vz 1
Va 1
Va 1
( ) ( ) ( ) =
Vg Vc Vk
Vz 1 Va 1 Va 1
=( ) ( ) ( )
Vg Vc Vk
1

=( )

1
=1

1+1
1+1
1+1 = 1




=1


=

1
=

Now putting value of in equation (1)

1
[ ( 1) + c 1 1 ]
= 1
c 1 [ 1 + 1 ]
8

Matlab Code:

function [eff, wc, wt]=turbodual(r, p1, t1, gam, q1, q2, t6)
% OUTPUT:
% eff= Efficiency of impulse turbocharged dual cycle
% wc= Compressor work in Kj/Kg
% wt= Turbine work in Kj/Kg
% INPUT
% r= compression ratio
% p1= Pressure at the beginning of compression (bar)
% t1= Temperature at the beginning of compression (bar)
% gam= 1.4
% q1= Amount of heat added at constant volume (Kj/Kg)
% q2= Amount of heat added at constant pressure (Kj/Kg)
% t6= Exhaust gas temperature (K)
% REFERENCE
% INPUT VALUE
% 1) Introduction to internal combustion engines by Richard Stone
% Page No. 325-360
% 2) Internal Combustion Engine Fundamentals by John B. Heywood
% Page No. 248-270
% 3) Internal Combustion Engine by V Ganeshan
% Page no. 665-694

p2= (p1*(r ^ gam ));


fprintf ('\n Value of p2=%0.2f',p2);
t2= (t1*(r)^(gam-1));
fprintf ('\n Value of t2=%0.2f',t2);
v1= (0.00287*t1)/p1;
fprintf ('\n Value of v1=%0.2f',v1);
v2= v1/r;
fprintf ('\n Value of v2=%0.2f',v2);
t3= (q1/0.717) + (t2);
fprintf ('\n Value of t3=%0.2f',t3);
p3= (p2*t3) / t2;
fprintf ('\n Value of p3=%0.2f',p3);
v3=v2;
fprintf ('\n Value of v3=%0.2f',v3);
t4= (q2/1.004) + t3;
fprintf ('\n Value of t4=%0.2f',t4);
v4= (t4*v3) / t3;
fprintf ('\n Value of v4=%0.2f',v4);
p4= p3;
v5= v1;
t5= t4 * ((v4/v5) ^ (gam-1));
fprintf ('\n Value of t5=%0.2f',t5);
p5= p4* (v4/v5) ^ gam;
v6= ((t5*(v5) ^ (gam-1)) / t6) ^ (1/ (gam-1));
fprintf ('\n Value of v6=%0.2f',v6);
p6= ((p3)*(v4/v6) ^ (gam));
fprintf ('\n Value of p6=%0.2f',p6);
lambda= p3/p2;
% lambda= Pressure ratio
rho = v4/v3;
% rho= cut-off ratio
fprintf ('\n Value of rho=%0.2f',rho);
9

w= 1.4;
phi= (rho*(lambda ^ (1/gam)))/w;
% phi= Coefficient of cooling in pre-cooler
fprintf ('\n Value of phi=%0.2f',phi);
v7= v6/phi;
fprintf ('\n Value of v7=%0.2f',v7);
p7= p6;
t6= (t4*(v4/v6) ^ (gam-1));
fprintf ('\n Value of t6=%0.2f',t6);
t7= (p7*v7)/0.00287;
fprintf ('\n Value of v7=%0.2f',v7);
v8= v7/w;
fprintf ('\n Value of v8=%0.2f',v8);
p8= p7 * (v7/v8) ^ (gam);
fprintf ('\n Value of p8=%0.2f',p8);
t8=t7*w;
fprintf ('\n Value of t8=%0.2f',t8);
wc= (1.01*(t8-t7));
% wc=compressor work in Kj/Kg;
% Cp air= 1.01 Kj/Kg;
% reference
% Introduction to internal combustion engines by Richard Stone
% Page No. 353
fprintf ('\n Value of compressor work=%0.2f',wc);
wt= (1.15*(t5-t6));
% wt= turbine work in Kj/Kg;
% Cp ex= 1.15 Kj/Kg;
% reference
% Introduction to internal combustion engines by Richard Stone
% Page No. 353
fprintf ('\n Value of turbine work=%0.2f',wt);
v12 = linspace (v2, v1, 200);
p12 = (p1*(v1.^gam)) ./ (v12.^gam);
subplot (2, 2,1);
plot (v12, p12); hold on
p23 = linspace (p2, p3, 400);
plot (v2, p23, 'r');
v34 = linspace (v3, v4, 200);
plot (v34, p3, 'g');
v45 = linspace (v5, v4, 200);
p45 = (p4*(v4 .^ gam)) ./ (v45 .^ gam);
plot(v45, p45,'black');
p15 = linspace (p1, p5, 400);
plot (v1, p15, 'r');
xlabel ('volume');
ylabel ('pressure');
legend ('graph of dual cycle');

subplot (2, 2, 2);


v56 = linspace (v6, v5, 200);
p56 = (p5*(v5 .^ gam)) ./ (v56 .^ gam);
plot(v56, p56); hold on
v67 = linspace (v6, v7, 200);
plot (v67, p6, 'red');
legend ('expansion in turbine');
v78 = linspace (v7, v8, 200);
p78 = (p7*(v7 .^ gam)) ./ (v78 .^ gam);plot (v78, p78, 'b');
10

v18 = linspace (v1, v8, 200);


plot(v18, p8);
xlabel ('volume');
ylabel ('pressure');
rc = v7 / v8;
rt = rc * r;
eff= 1 - (gam / (rt ^ (gam-1))) * (((rc ^ (gam-1)) + (rho*lambda ^ (1/gam)) -
w) / (lambda - 1 + gam*lambda*(rho-1)));
fprintf ('\n Value of efficiency of turbocharged dual cycle =%0.2f',eff);
end

Steps to run the program:


1. Open the MATLAB interface.
2. In command window type turbodual (14, 2.5, 300, 1.4, 400, 400, 600).
3. Press the ENTER and wait for the results.

Sample Input:
turbodual (14, 2.5, 300, 1.4, 400, 400, 600)

Output :
turbodual (14, 2.5, 300, 1.4, 400, 400, 600)

Value of p2=100.58
Value of t2=862.13
Value of v1=0.34
Value of v2=0.02
Value of t3=1420.01
Value of p3=165.67
Value of v3=0.02
Value of t4=1818.42
Value of v4=0.03
Value of t5=698.56
Value of v6=0.50
Value of p6=3.42
Value of rho=1.28
Value of phi=1.31
Value of v7=0.39
Value of t6=600.00
Value of v7=0.39
11

Value of v8=0.28
Value of p8=5.48
Value of t8=642.99
Value of compressor work=185.55
Value of turbine work=113.34
Value of efficiency of turbocharged dual cycle =0.4824

Fig.1 Turbocharged Dual Cycle

REFERENCES:

1) Introduction to internal combustion engines


by Richard Stone
Page No. 325-360

2) Internal Combustion Engine Fundamentals by


John B. Heywood
Page No. 248-270
HEAT TRANSFER IN ENGINES

Importance of Heat transfer in engine:


There is huge amount of heat generated in the engine due to combustion of fuel. Due to
this gases in the cylinder are heated. The maximum gas temperature in the cylinder of an engine
is of the order of 2500K. But for various reasons and metallurgical limitations metal temperature
must be less than 400C (673K) for cast iron and 300C (573K) for aluminum alloys. The gas
side surface of the cylinder wall must be kept below 180C to prevent deterioration of lubricating
oil film. Critical areas like piston crown, exhaust valve, Spark plug must be kept cool to avoid
knock and preignition problems which result from overheated spark plug electrodes or exhaust
valves.

MODES OF HEAT TRANSFER :-

1. Conduction:-
Heat is transferred by molecular motion, through solids and through fluids at
rest, due to a temperature gradient. The heat transfer by conduction, per unit area per unit time, q,
in a steady situation is given by Fourier's law.
=
Heat is transferred by conduction through the cylinder head, cylinder walls, and
piston; through the piston rings to the cylinder wall; through the engine block and manifolds.

2. Convection :-

1
Heat is transferred through fluids in motion and between a fluid and solid surface in
relative motion. When the motion is produced by forces other than gravity, the term forced
convection is used. In engines the fluid motions are turbulent.
Heat is transferred by forced convection between the in-cylinder gases an the cylinder
head, valves, cylinder walls, and piston during induction, compression, expansion, and exhaust
processes.
q hc (T TW ) ........hc is heat transfer coefficient

cp
m n
L
cons tan t
hc L
(1)
k k

i.e, Nu C Rem Pr n

where L and v are a characteristic length and velocity.


For gases, the Prandtl number (c,p/k) varies little and is about 0.7.

3. Radiation :-
Heat transfer by radiation occurs from the high-temperature combustion gases and
the flame region to the combustion chamber walls (although the magnitude of this radiation heat
transfer relative to convective heat transfer is only significant in diesel engines). Heat transfer by
radiation to the environment occurs from all the hot external surfaces of the engine.
The heat flux from one plane black body at temperature T, to another at
temperature T2 parallel to it across a space containing no absorbing material is given by
q (T14 T24 )

where is the Stefan-Boltzmann constant 5.67 x 10-8W/m2K4


In real black bodies do not exist. So we multiply equation by emissivity.
Heat transfer in engine:-

2
Fig.1. Temperature distribution and heat flow across the combustion chamber wall.
For a steady one-dimensional heat flow through a wall as indicated in Fig.1, the
following equations relate the heat flux q = Q/A and the temperatures indicated :

Gas Side: q qcv q R hc ,g (Tg Tw,g ) (Tg4 Tw4.g )

k (Tw,g Tw,c )
Wall: q qCN
tw

coolant side: q qCV hc ,c (Tw,c Tc )

INSTANTANEOUS HEAT TRANSFER CORELATIONS


Instantaneous heat transfer correlations are used to find heat transfer coefficients, heat
flux at each crank angle of engine. Various models have been evolved over time to find
instantaneous heat transfer. the most basic equation used for calculation is
hL
Nu C Re a Pr b
k
(2)

3
Woschni's Heat transfer model
Probably the most widely used approach of this category is the one suggested by
Woschni [2] who assumes an analogy between the cylinder flow and a turbulent flow through a
circular tube.
Consequently, the cylinder length L is the cylinder bore B and the constant a and b are
chosen to be 0.8 and 0.4, respectively. Replacing the Reynolds number by

cB p cB p
Re RT
RT (3)
Where c is the characteristic flow velocity,
Eq. 2 now becomes
hB
Nu C Re 0.8 Pr 0.4
k
0.8
p cB
C Pr 0.4
RT (4)
Introducing the property relations
x y
k T T
,
k o T0 0 T0
Pr= const. 0.74
and setting the characteristic flow velocity to the mean piston velocity cm, we can solve for the
heat transfer coefficient h:
h C * B 0.2 p 0.8 c m T r ;
0.8
r = 0.8 (1+ y ) - x (5)

By comparison with experimental data the exponent r of the temperature dependency is


found to be r = 0.53 and the constant C* becomes 3.26 W/(m2K). However, the above relation is
applicable only for motored engine operation. Since the combustion process during red
operation significantly increases the turbulence level and thus the heat transfer coefficient, Eq. 5
has to be corrected by replacing cm by a velocity term v, that takes into account both the mean
piston velocity and the combustion induced turbulence:

h = 3.26B-0.2 p0.8 v0.8 T-0.53 (6)

where

4
B is in m ,p is in kPa , T is in K ,v is in m/s

Where
Vd T1
v C1 cm C2 ( p pmot ) (7)
p1V1
index 1 refers to the state at the start of compression, i.e. at IVC, and Vd is the displacement
volume. The difference in cylinder pressure between red and motored engine operation (p-pmot)
accounts for the turbulence increase by combustion and the constants C1 and C2 are suggested to
be:

cu
6.18 0.417 c : gas exchange

C1
m
(8)
2.28 0.308 cu : compression/ expansion
cm

6.22 10 3 m /( sk ) : prechamber engine


C2 3
(9)
3.24 10 m /( sk ) : direct injection engine

Equation 8 has been validated for intake swirl ratios, cu/cm, between 0 and 3.
Annands Heat transfer coefficient model:
Annands and Woschnis heat transfer models differed in the fact that Annands approach
separated the convective and radiative terms. Annands method solved for the heat transfer
coefficient by assuming pipe-like fluid dynamics, and using the in-cylinder density, and
Reynolds and Nusselt numbers as functions of time.
Using Annands method, Newtons law of cooling can be broken into convective and
radiative terms as follows:
Qw (hc hr ) Aw (T Tw )dT (12)

where hc is the convective heat transfer coefficient and hr is the radiative heat transfer
coefficient.
The convective heat transfer coefficient can be extracted from the relationship between
the Nusselt number and fluid properties:

5
k gas(W / mK ) Nu
h(W / m2 K )
B( m) (13)

where kgas is the gas thermal conductivity, Nu is the Nusselt number, and B is the cylinder bore.
With an iterative solver, the thermal conductivity of the cylinder gas can be modeled
using a polynomial curve-fitting of experimental data. Blair suggests using the curve-fitted
equation[4]:

W 3 5 8 2
k gas 6.1944 10 7.3814 10 T ( K ) 1.2491 10 T ( K )
mk (14)
where is the instantaneous cylinder temperature, and the thermal conductivity is output in units
of [W/mk] .
The Nusselt number can be described relative to the Reynolds number and the type of
engine:
Nu a Re0.7 (15)

where is a constant having a value of 0.26 for a two-stroke engine and 0.49 for a four-stroke
engine, and Re is the instantaneous Reynolds number.
The Reynolds number is expressed as:
gas S p B
Re
gas
(16)
Where gas is the instantaneous cylinder gas density, S p is the mean piston velocity, and B is the
bore diameter and gas is instantaneous gas viscosity.

Since the model assumes ideal gas behavior, the cylinder gas density can be found by
rearranging the ideal gas law:
P( Pa )
gas
RairT ( K ) (17)
Where Rair is the fluid-specific gas constant, and an assumed value of 287[J/Kg*K ] was used for
this variable.
As with the thermal conductivity, the cylinder gas viscosity was modeled using empirical
equations. According to Blair, the cylinder gas viscosity can be expressed as:

6
kg
gas 6 8 12 2
7.457 10 4.1547 10 T ( K ) 7.4793 10 T K
ms (18)

where the instantaneous cylinder temperature must be provided in units of [K] , and the gas
viscosity is output in units of [Kg/ms] .
Although the radioactive heat transfer coefficient is small, it was decided that radiation
should be included in considering overall heat losses in the model. The radiative heat transfer
coefficient is defined as :

W 9 T T w
4 4
hr 2 4.25 10 0.75
m k T Tw (19)

where the instantaneous cylinder temperature and wall temperature must be provided in units of
[K]. With known pressure and temperature traces from the single-zone calculations, Annands
method could then be used to calculate heat losses.

Hohenberg's Heat transfer model

Hohenberg [3] proposed an in-cylinder heat transfer equation by modifying the Woschni
equation in the DI diesel engines with swirl. Equation (20) describes the in-cylinder HTC
proposed by Hohenberg:

h 3.26 P 0.8 T 0.4 V 0.06 ( S p 1.4)0.8 (20)

7
REFERENCES

[1] Annand WJ (1963) Heat Transfer in Cylinder of Reciprocating Internal Combustion Engines.
Proc Inst Mech Engineers, vol 177, No 36, pp 973-990

[2] Woschni G (1967) A Universally Applicable Equation for the Instantaneous Heat Transfer
Coefficient in Internal Combustion Engine. SAE PAPER 670931

[3] Hohenberg G (1979) Advanced Approaches for Heat transfer Calculation. SAE Paper
790825

[4]G. P. Blair, Design and Simulation of Four Stroke Engines [R-186]. Society of Automotive
Engineers Inc, 1999. (for Annand)

8
Matlab Program

1)Input required:- htc ( RPM, L, B, l, CR, theta_b, theta_0)


where L=stroke(m), B=Bore(m), l= connecting rod length(m), theta_b= burn duration(degree)
theta_0= start of combustion(degree)......compression starts at 0 degree

2)Reference values RPM=2500, L=B=0.1m, l=0.15, theta_b= 60, theta_0= 160

3)Output: plots comparing woschni and aanands' heat transfer cofficient, heat flux, heat
transfered.

PROGRAM:-

function htc(RPM,L,B,l,CR,theta_b,theta_0)
clc;
%_________________________________________________________________________

eta_f = theta_0+theta_b; %Final Comb. Angle [degrees]


IVC = 0; %Time [degrees] when Intake Valve Closes
EVO = 360; %Time [degrees] when Exhaust Valve Opens
%_________________________________________________________________________
%Engine Calculations Based On Previous Inputs
%Assumes Average Surface Area In Which Heat Transfer Occurs
A_p = (pi/4)*B^2; %Cross Sectional Piston Area [m^2]
A_ch = 2*A_p; %Cylinder Head Surface Area (in chamber)
V_d = A_p*L; %Displaced Volume Of Engine [m^3]
N = RPM/60; %Co=nverts RPM to RPS [1/s]
S_bar_p = 2*L*N; %Calculates Mean Piston Speed [m/s]
a = L/2; %Calculates Crank Radius (1/2 stroke)[m]
V_TDC = (V_d/(CR-1)); %Calculates Clearance Volume [m^3]
V_BDC = V_d+V_TDC; %Cyl. Volume At BDC [m^3]
%_________________________________________________________________________

%Initial Preallocation Of Matrices (Second Preallocation In Loops Needs To


%Be Included (Do Not Delete)

9
V(1:360)=zeros;DV(1:360)=zeros;rho(1:360)=zeros;mu(1:360)=zeros;
kg(1:360)=zeros;hr(1:360)=zeros;X(1:360)=zeros;M_F(1:360)=zeros;
DX(1:360)=zeros;Re(1:360)=zeros;Nus(1:360)=zeros;hg_a(1:360)=zeros;
DQ(1:360)=zeros;Q(1:360)=zeros;DT(1:360)=zeros;
DP(1:360)=zeros;P(1:360)=zeros;T(1:360)=zeros;
s(1:360)=zeros; Pm(1:360)=zeros; hg_w(1:360)=zeros;U(1:360)=zeros;
Qa(1:360)=zeros; Qw(1:360)=zeros; qa(1:360)=zeros;qw(1:360)=zeros;
Qh(1:360)=zeros; Qe(1:360)=zeros; qh(1:360)=zeros;qe(1:360)=zeros;
hg_h(1:360)=zeros ; hg_e(1:360)=zeros ;A(1:360)=zeros ;
%_________________________________________________________________________
%Fuel Inputs/Efficiencies
AF_ratio_stoich = 15.09; %Gravimetric Air Fuel Ratio (Stoich)
AF_ratio_mol_sotich=14.7; %Molar Air_Fuel Ratio (Stoich)
lambda = 1; %Excess Air Coefficient
AF_ratio_ac = lambda*AF_ratio_stoich; %Actual Air Fuel Ratio
AF_ratio_mol=lambda*AF_ratio_mol_sotich;
LHV = 44.6e6; %Lower Heating Value Of Fuel Mixture [J/kg]
eta_combmax = 1; %Assumed MAX COmb. Efficiency
%Predicts Combustion Efficiency (Reference To Blair)
eta_comb=1; %eta_combmax*(-1.6082+4.6509*lambda-2.0764*lambda^2);
%Atmospheric Inputs
P_atm = 101325;
T_atm = 300;
P_BDC = P_atm; %Inlet Pressure[Pa] Moscow,ID
R_air = 287; %Gas Constant For Air [J/kg-K]
gamma(1:360) = 1.4; %Preallocate Gamma Array (sets initial value)
Tw =400; %Assumed Wall Temperature
%________________________________________________________________________

for k = 1:2

T_BDC = T_atm; %Assumed Inlet Temperature [K]

%Calculate Mass of Air In Cylinder/ Mass Of Fuel Based On AFR


rho_a = P_BDC/(R_air*T_BDC); %Air Density kg/m^3
m_a = rho_a*V_d; %Mass of Air In Cylinder [kg]
m_f = m_a/AF_ratio_ac; %Mass Of Fuel In Cylinder [kg]

10
m_c = m_a+m_f; %Mass In Cylinder
%Specifying Initial Conditions For Loops
%DV,DX,etc. Are Relative To Change In Theta (i.e. DV/Dtheta)
theta(1:360)=zeros; %Starting Crank Angle [deg]
V(1:360)=zeros; %Preallocate Volume Array
V(1)=V_BDC; %Starting Combustion Chamber Volume [m^3]
DV(1:360) = zeros; %Preallocate Change In Volume Array
DV(1) = 0; %Specifying Initial Change In Volume [m^3}
P(1:360)=0; %Preallocate Pressure Array
P(1)=P_BDC;
DP(1:360) = zeros; %Specifying Initial Change In Pressure
T(1:360)=zeros; %Preallocate Temperature Array
T(1) = T_BDC; %Inlet Temperature [K]
DT(1:360) = zeros; %Specifying Initial Change In Temperature
gamma(1)=1.4; %Initial Gamma Input
X(1:360) = 0; %Preallocate Mass Burn Array
DX(1:360) = zeros; %Preallocate Change In Mass Burn Fraction [unitless]
DQ(1:360) = zeros; %Preallocate Heat Release Array
Q(1:360)=zeros; %Preallocate Heat Array
M_F(1:360) = zeros; %Preallocate Mass In Comubstion Chamber Array
rho(1:360) = zeros; %Preallocates Ideal Gas Law array
rho(1) = P(1)/(R_air*T(1)); %Initial Value Ideal Gas Array
mu(1:360)=zeros; %Preallocate Viscosity Array
mu(1)=7.457*10^(-6) + 4.1547*10^(-8)*T_BDC - 7.4793*10^(-12)*T_BDC^(2);
kg(1:360)=zeros; %Preallocate Thermal COnductivity Array
kg(1) = 6.1944*10^(-3) + 7.3814*10^(-5)*T_BDC - 1.2491*10^(-8)*T_BDC^(2);
hr(1:360) = zeros; %Preallocate Radiation Coefficient Array
hr(1) = 4.25*10^(-09)*((T(1)^4-Tw^4)/(T(1)-Tw)); %Initial Rad. Coeff
Re(1:360)=zeros; %Preallocate Reynolds Value Array
Re(1)=rho(1)*S_bar_p*B/mu(1); %Initial Reynolds Value
Nus(1:360)=zeros; %Preallocating Nusselt Number Array
Nus(1)=.49*Re(1)^(.7); %Initial Nusselt Number
hg_a(1:360)=zeros; %Preallocate Heat Transfer Coefficient Array
hg_a(1)=kg(1)*Nus(1)/B; %Initial Heat Transfer Coefficient
s(1:360)=zeros; %Preallocates Distance Crank/Piston Axes Array
s(1) = -a*cosd(theta(1))+sqrt(l^2 - a^2*sind(theta(1))^2);%Initial Val.

end

11
theta=1:360;

for i = 2:360
%Specifies Distance Between Crank/Piston Axes As A Function Of theta
s(i) = -a*cosd(theta(i))+sqrt(l^2 - a^2*sind(theta(i))^2);
%Specifies Volume As A Function Of Crank Angle
V(i) = V_TDC +((pi/4)*B^2)*(l + a - s(i));
%Specifies Change In Volume As A Function Of Crank Angle
DV(i) = V(i)-V(i-1);
%Calculates Density As A Function Of Crank Angle
rho(i) = P(i-1)/(R_air*T(i-1));
%Calculates Viscosity As A Function Of Temperature
mu(i) = 7.457*10^(-6) +4.1547*10^(-8)*T(i-1) - 7.4793*10^(-12)*T(i-1)^(2);
%Calculating Instantaneous Thermal Conductivity of Cylinder Gas
kg(i) = 6.1944*10^(-3) + 7.3814*10^(-5)*T(i-1) - 1.2491*10^(-8)*T(i-1)^(2);
%Calculating The Radiation Heat Transfer Coefficient
hr(i) = 4.25*10^(-09)*(((T(i-1)^4)-(Tw^4))/(T(i-1)-Tw));
%Instantaneous Suface Area (For Heat Transfer)
A(i) = A_ch + A_p + pi*B*(l+a-s(i));

%_____________________________________________________________________
%Specifies Mass Fraction Burn As A Function Of Crank Angle (Weibe Fcn.)
%Also Specifies Mass Of Fuel In Combustion Chamber As A Function Of
%Theta
if theta(i)<theta_0
X(i)=0;
else
X(i) = 1 - exp(-5*((theta(i)-theta_0)/theta_b)^3);
if theta(i) < theta_f
M_F(i) = V(theta_0-1)*rho(theta_0-1)/(lambda*AF_ratio_mol_sotich);
end
end

%____________________________________________________________________
%Specifies Change In Mass Fraction Burn As A Function Of Crank Angle
DX(i) = X(i) - X(i-1);

12
%____________________________________________________________________
%Incorporating The Annand Method To Predict Heat Transfer
%Calculating Reynolds Number
Re(i)=rho(i)*S_bar_p*B/mu(i);
%Calculating Nusselt Number (constant=.26 two stroke, .49 4 stroke)
Nus(i)=.49*Re(i)^(.7);
%Calculating Heat Transfer Coefficient Using Annand Method
hg_a(i)=kg(i)*Nus(i)/B;
%Calculates Convective Losses Into Wall As A Function Of Crank Angle
%Calculates Change In Heat Transfer (total) As A Function Of Crank Angle
Q(i) = M_F(i)*LHV*X(i);
%____________________________________________________________________
%Specifies Pressure and Temperature Increases Between Intake Valve
%Closing and Exhaust Valve Opening
if IVC< theta(i)
DT(i)=T(i-1)*(gamma(i-1)-1)*((1/(P(i-1)*V(i-1)))*DQ(i)-(1/V(i-1))*DV(i));
DP(i)=(-P(i-1)/V(i-1))*DV(i)+(P(i-1)/T(i-1))*DT(i);
P(i) = P(i-1)+DP(i);
end
if EVO < theta(i)
P(i) = P_atm;
end
if 200 < theta(i)
if P(i)<=P_atm
P(i)=P_atm;
if P(i)<= Pm
P(i) = Pm;
end
end
end
%Returns Temperature Values To Beginning Of Loop
T(i) = T(i-1)+DT(i);

%----------------- woschni correlation ----------------%

Pm(i)= P_atm*(V_BDC/V(i))^1.4; % motoring pressure

13
% Calculation of Average gas velocity
if theta_0<=theta(i)<=theta_f
U(i) = 2.28*S_bar_p + 0.00324*T_BDC*V_d*(P(i)-Pm(i))/(V_BDC*P_BDC);
else
U(i) = 2.28* S_bar_p;
end

% Following relation gives instantaneous Heat transfer coefficient by


% woschni corelation
hg_w(i) = 3.26*((P(i)/1000)^0.8)*(U(i)^0.8)*(B^-0.2).*(T(i)^-0.55); %as P in
KPa

%--------- Calculation Heat transfer and heat flux --------------%

Qa(i) =(hg_a(i)+hr(i))*A(i)*(T(i)-Tw)/1000; %Annand's heat transfer


qa(i)= Qa(i)/A(i); %Annand's heat flux

Qw(i) =hg_w(i)*A(i)*(T(i)-Tw)/1000;% woschni's heat transfer


qw(i)= Qw(i)/A(i); %woschni's heat flux

end

%_________________ Ploting Statements ____________________________%


plot(theta, hg_a,theta, hg_w)
title('Heat transfer coeff. Plot');
xlabel('crank angle [degrees]');
ylabel('Heat transfer coeff. (W/m2K)');
legend('Annand','Woschni');
grid on;

figure(2)
plot(theta, qa,theta,qw)
title('Heat Flux Plot');
xlabel('crank angle [degrees]');

14
ylabel('Heat Flux q (KW/m2)');
legend('Annand','Woschni');
grid on;

figure(3)
plot(theta, Qa,theta,Qw)
title('Heat transfer Plot');
xlabel('crank angle [degrees]');
ylabel('Heat transfer Q(KW)');
legend('Annand','Woschni');
grid on;

Outputs

1) Heat Transfer Co-efficient:-

15
2) Heat Flux:-

3)Heat Lost:-

16
1

The influence of engine speed and load on the heat transfer between gases and in-
cylinder walls at fired and motored conditions of an IDI diesel engine.

Introduction
Heat transfer through the cylinder side walls is an important process in determining
overall performance, size and cooling capacity of an internal combustion engine (ICE). It affects
the indicated efficiency because it reduces the cylinder temperature and pressure, and there by
decreasing the work transferred on the piston per cycle. The heat loss (transfer) through the walls
is in the range of 1015% of the total fuel energy supplied to the engine during one working
cycle.
From the in-cylinder point of view, the heat transfer changes the local and instantaneous
temperatures that are of a critical importance in controlling NOx emission formation. High
temperatures lead to thermal stresses of material, and impact on fatigue failure limits of various
engine components, thus causing fatigue cracking or deforming cylinder bore dimension and
valve stems. Kept below certain limits of the combustion chamber side wall temperature is
necessary to prevent the lubricating oil film from deterioration and its viscosity from
diminishing. In order to avoid from pre-ignition and knock risks resulting from overheated spark-
plug electrodes and exhaust valves in spark-ignition (SI) engines, the spark-plug and valves must
be kept cool. In addition to these, heat transfer to the incoming air reduces volumetric efficiency.
There are a number of studies pertaining to the effect son heat transfer characteristics of
the various engine operating parameters. Alkidas experimentally compared the variations of the
heat flux by using thermocouples mounted at different locations of combustion side of cylinder
head of a single-cylinder SI engine at the motored and fired conditions. He showed that the
magnitude of peak heat fluxes in the combustion chamber increased with the engine speed sat
both conditions. He also compared the heat flux values obtained by Woschnis equation with the
experimentally measured ones and observed that the ones obtained by thatequation were in good
agreement with the experimental results. Watts et al. [4] showed that, at constant load for SI
engines, the average heat transfer through the cylinder walls as a percentage of total fuel energy
diminished almost linearly with increasing engine speed first and thereafter did not change at
higher engine speeds. By using Woschni and Annand equations, Karamangil et al. [5]
investigated parametrically how the convective heat transfer coefficients for gas side and coolant
side varied with different engine operating parameters, such as the engine speed, compression
ratio, excess air coefficient, combustion duration, inlet pressure and temperature. Rakopoulos et
al. [6,7] carried out the heat transfer analysis at some engines under various engine operating
parameters. In [6], it was examined the heat release rates and the heat flux variations for both
main chamber and pre-chamber of a turbocharged IDI diesel engine at the different load ranges.
In [7], they studied the instantaneous in-cylinder and exhaust pipe heat fluxes under different
speed and load ranges of a diesel engine with air-cooled and one-cylinder. Shiling et al. [8]
compared the measured instantaneous heat transfer coefficients with the calculated values and
found that the calculated results were in a good agreement with the values obtained
experimentally.
As can be seen in the relevant literature, few heat transfer studies for IDI engines seem to
have been performed recently. Present work, intended for a contribution to the open literature,
deals with how to vary the heat transfer coefficients and heat fluxes with the crank angle at
2

different speed and load ranges for fired and motored operations by using the Woschni
expression.

Heat transfer calculation between gases and in cylinder walls


Heat transfer phenomenon in the ICEs is a highly comprehensive and complex issue
because of being transient, three dimensional, and the effect of periodic pressure and temperature
fluctuations of the charging. Furthermore, when radiation heat transfer as well as convection heat
transfer in cylinder is considered, the issue becomes more complicated. The heat transfer by
radiation, taking place from the high temperature solid soot particulates during combustion
process, has a dominant effect in diesel engines and ranges about from 10% to 40% of total heat
loss to the walls, whereas this ratio is generally negligible and just about 510% in SI engines.
Fig. illustrates both the convective and radioactive components, and the heat transfer
between gas and in-cylinder wall, and conduction through the combustion chamber wall, and
convection to the coolant. Mean gas and coolant temperatures, Tmg and Tmc, are depicted with
dotted lines.
Applying the assumption of steady heat transfer for the present study, the heat flux can be
calculated from the Newtons cooling law by multiplying the simultaneous heat transfer
coefficient with the temperature difference, i.e.
q( ) Q( ) A h( )(Tg ( ) Tw )
Since the last 45 years, many models with global validity in relation to the computing of the in-
cylinder heat transfer coefficient have been proposed by a number of investigators, e.g. Woschni,
Annandand Hohenbergand so on, whose models have relied on dimensional analysis for
turbulent flow that correlates the Nusselt, Reynoldsand Prandtl numbers, which is
Nu a Re b Pr n
Where Pr has been omitted because of changing little ingases, so its effect can be included in the
a coefficient.
In particular, the Woschni correlation has frequently been used in the heat transfer studies with
proper constants in todays SI and diesel engines, and also it has been correlated conveniently
even for HCCI engines recently

Fig. Schematic of the overall heat transfer process in the cylinder

Instantaneous heat transfer coefficient adopted from Woschni is calculated by,


h( ) 3.26P( ) b Tg ( ) 0.751.62b D b1 w( ) b
3

He assumed the b exponent is 0.8 and emphasized that effective gas velocity, w, consists of
two contributions. The first contribution is scaled with mean piston motion and swirl, and the
second contribution is related to turbulence effects and DP pressure rise resulted from
combustion, and of course this contribution includes the influence of radiation. The term
TV
w( ) (2.28 0.308vu Up)Up C R s P
PuVR
for the compression and expansion strokes
w( ) (2.28 0.308vu Up)Up
for scavenging period, namely the induction and exhaust strokes. Where vu is defined as
DN an
vu
60
and Nan can be calculated from 0.7B below the cylinder head in a steady flow test, Up is mean
piston speed, Vs is stroke volume, TR, PR and VR are respectively taken as the temperature and
pressure of working gases and total volume on piston at a chosen reference state, either at intake
valve closing (IVC) or at the start of combustion, her IVC. DP, P()-Pm() is instantaneous
difference in pressure between the firing and motoring engine at the same crank angle. In the
present study, the motoring pressure Pm was measured experimentally instead of a calculation
method suggested by Watson and Janota and was observed the different results than that of the
model at the two different engine speeds as indicated in Fig. 2. At the start of combustion, C
coefficient has been correlated by Woschni as 3.24 x 10-3 for both the diesel engines and SI
engines, whereas as 6.22 x 10-3 for IDI diesel engines because of higher turbulence effects.
The other approach for the instantaneous heat transfer coefficient, developed by Annand,
can be expressed as
k ( ) b
h( ) a Re
D
The value of a constant, which represents the level of convective heat transfer, varies with
intensity of charging motion and combustion chamber design. At normal combustion, it varies
from 0.25 to 0.8 and increases directly with increasing the intensity of charging motion. The
index b varies from 0.7 to 0.8. k thermal conductivity of the gas as a function of Tg can be
calculated from following formula;
k ( ) 3.17 10 4 Tg ( ) 0.772
or from
C p ( )
k ( )
Pr
For above eq, k in the present work was used. Cp is constant pressure specific heat of cylinder
gases. , viscosity of combustion products, is a function of temperature and equivalence ratio,
and almost independent of pressure. It is computed from following equation:
( ) 3.3 10 7 Tg ( ) 0.7 (1 0.027 )
Pr, Prandtl dimensionless number, is nearly constant for ordinary gases and equal to
0.74.However, we have calculated it from following equation due to being measured 1
Pr 0.05 4.2( 1) 6.7( 1) 2
4

For the specific heats ratio , one generally takes it between 1.3 and 1.35 values. Here, was
found from equivalence ratio ( = -1) and this variable has been continuously updated during
the different engine operating conditions.
1.4 0.16
Reynolds number is formed with a characteristic speed equal to the Up mean piston velocity, a
characteristic length equal to the D cylinder bore, gas density and dynamic viscosity as
follows:
( )UpD
Re( )
( )
The density of cylinder contents during time between IVC and EVO is mainly a function of
instantaneous cylinder volume
m
( )
V ( )
Assuming perfect gas behavior,
PVM mix
m
Ru Tg
Mmix is the molar mass of the mixture of air and fuel. The instantaneous cylinder volume and
area as a function of crank angle are given by
D 2
V ( ) Vc x( )
4
x() represents for distance from TDC, and thus according to the crank angle

x( ) l R R cos( ) (l 2 R 2 sin 2 ( ))
Hohenberg made some modifications in the Woschnis formula, such as the instantaneous
cylinder volume instead of bore, the effective gas velocity, and the exponent of the instantaneous
gas temperature. Hohenbergs equation therefore is described as
h( ) 3.26P( ) 0.8 Tg ( ) 0.4 V ( ) 0.06 (Up c) 0.8
where c is the calibration constant that Hohenberg suggested to be 1.4 for the engine he studied.
As the cylinder gas temperature, the gases far away from the wall are used. Hence the
instantaneous gas temperature was calculated by assuming the first law of thermodynamics and
using the cylinder pressure;
1
P( )
Tg ( ) Tin
Pin
can be computed as a function of equivalence ratio by means of Eq. (12) as cited already for
hydrocarbon fuels.
5

Matlab Program:
1)Steps for running the programe
Put all files in single folder
Open the ptheatha.m file
Run
You get the output

functionPcal=N1000(Pthetacal)
Ptheta1000=[-180.397 -174.046 -167.696 -161.345 -154.598 -146.262 -137.927 -
129.592 -120.463 -112.525 -101.808 -93.473 -84.741 -74.818 -65.292 -56.957 -49.416 -
41.874 -36.318 -31.951 -28.379 -24.41 -21.235 -18.853 -17.266 -15.281 -12.9 -11.312 -8.931 -
6.946 -4.961 -2.977 0.992 5.755 8.534 9.724 11.312 13.297 14.884 16.075 17.663
18.456 20.838 23.219 25.204 27.585 30.364 32.745 37.508 41.08 46.24 52.988 60.926
68.071 77.2 85.932 96.648 105.38 114.509 123.638 132.767 141.499 150.628 159.361
168.093 175.237 180];
P1000=[105.86 105.86 105.86 105.86 119.093 119.093 119.093 119.093 145.558 132.325
158.79 185.255 211.72 251.418 304.348 410.208 542.533 727.788 939.509 1151.229
1415.879 1706.994 1984.877 2236.295 2500.945 2739.13 3043.478 3361.059
3652.174 3916.824 4155.009 4340.265 4499.055 4432.892 4247.637 4062.382
3850.662 3612.476 3347.826 3149.338 2911.153 2712.665 2434.783 2209.83
1905.482 1693.762 1508.507 1270.321 1032.136 860.113 688.091 489.603 370.51
264.65 211.72 172.023 132.325 119.093 105.86 105.86 92.628 92.628 79.395 79.395
92.628 92.628 92.628];

Pcal=interp1(Ptheta1000,P1000,Pthetacal);
plot(Pthetacal,Pcal,'-r')
xlabel('Crank angle, deg')
ylabel('Pressure, kPa')
title('Comparison of the pressures from motored operation and Watson -Janota model')
end

functionPcal=N1500(Pthetacal)
Ptheta1500=[-179.603 -174.84 -167.696 -160.154 -153.407 -145.469 -137.927 -
131.974 -125.226 -116.494 -108.953 -99.824 -89.107 -79.581 -70.452 -59.338 -52.194 -
45.05 -39.89 -35.524 -31.158 -28.379 -25.601 -22.822 -20.838 -17.663 -15.678 -14.09 -12.106
-10.518 -8.931 -7.343 -5.358 -1.786 2.183 4.564 5.755 6.946 8.137 9.327 10.518
12.106 13.297 14.884 15.678 17.266 19.25 19.647 22.029 24.41 26.792 29.57 32.348
36.714 40.684 45.05 51.797 58.942 65.292 72.437 81.963 88.313 96.251 102.602 109.35
116.891 122.845 129.592 137.927 146.262 153.407 161.742 167.696 174.046 180.397]
P1500=[145.558 145.558 145.558 145.558 119.093 105.86 145.558 132.325 145.558 145.558
158.79 185.255 211.72 211.72 330.813 463.138 568.998 767.486 965.974 1217.391
1495.274 1759.924 2051.04 2355.388 2633.27 3030.246 3374.291 3665.406
3916.824 4207.94 4446.125 4671.078 4829.868 5028.355 4896.03 4724.008
6

4551.985 4366.73 4155.009 3930.057 3705.104 3480.151 3228.733 2950.851


2778.828 2580.34 2355.388 2143.667 1905.482 1640.832 1455.577 1243.856
1045.369 860.113 714.556 555.766 449.905 344.045 304.348 251.418 224.953 185.255 158.79
132.325 145.558 132.325 119.093 105.86 92.628 92.628 92.628 92.628 92.628 92.628
92.628]
Pcal=interp1(Ptheta1500,P1500,Pthetacal);
plot(Pthetacal,Pcal,'-k')
end

functionPcal=Nmodel(Pthetacal)
Pthetamodel=[-179.603 -174.443 -168.093 -160.551 -154.598 -147.85 -138.721 -
129.195 -121.257 -113.319 -104.587 -95.458 -85.932 -77.993 -70.055 -64.498 -58.942 -
51.797 -44.256 -40.287 -36.714 -33.936 -32.348 -29.967 -27.189 -25.204 -23.219 -21.632 -
20.441 -18.456 -16.869 -15.678 -14.487 -13.297 -11.709 -10.518 -9.327 -8.137 -6.549 -5.358 -
2.977 -0.992 2.58 5.358 6.152 7.74 8.534 9.724 10.915 11.709 12.503 13.693
14.884 16.075 17.266 18.853 20.044 21.632 23.219 24.807 27.585 28.776 31.555 34.333
37.508 41.477 45.843 51.797 60.132 67.674 72.834 80.772 88.71 95.061 102.999 110.937
118.082 124.035 129.195 133.958 143.881 150.232 157.376 164.123 169.283 179.603 179.603];
Pmodel=[119.093 119.093 119.093 119.093 119.093 119.093 119.093 119.093 119.093 145.558
132.325 185.255 224.953 264.65 304.348 383.743 449.905 595.463 793.951 939.509 1111.531
1283.554 1482.042 1654.064 1865.784 2103.97 2302.457 2514.178 2778.828
3017.013 3334.594 3612.476 3877.127 4168.242 4446.125 4737.24 5001.89
5319.471 5637.051 5875.236 6113.422 6258.979 6206.049 5954.631 5742.911
5491.493 5253.308 5068.053 4803.403 4525.52 4287.335 4075.614 3824.197
3599.244 3347.826 3109.641 2805.293 2567.108 2289.225 2064.272 1892.25
1654.064 1468.809 1283.554 1098.299 873.346 727.788 555.766 423.44 370.51
291.115 264.65 211.72 211.72 185.255 158.79 172.023 158.79 158.79 158.79 158.79
119.093 119.093 119.093 119.093 119.093 119.093];
plot(Pthetamodel,Pmodel,'-r')
end

clc
clearall
theata=-180:0.5:180;
N1000(theata)
holdon
N1500(theata)
Nmodel(theata)
7

2)
Steps for running the programe
Put all files in single folder
8

Open the wah.m file


Run
You get the output
clc
clearall
theata=-180:0.00005:180;
htcw(theata)
holdon
htca(theata)
htch(theata)

functionhcal=htcw(thetacal)
theta=[-179.995 -175.085 -170.176 -164.857 -157.494 -152.176 -
146.039 -139.084 -135.405 -130.497 -124.358 -119.04 -113.312
-106.357 -100.221 -93.266 -86.313 -78.54 -69.947 -62.174 -54.81 -
47.035 -41.308 -34.35 -28.619 -22.07 -17.976 -17.965 -14.683 -10.174 -8.932
-8.512 -8.087 -7.252 -6.829 -6.816 -6.392 -6.376 -5.953 -5.12 -
4.695 -4.27 -4.254 -3.422 -2.586 -1.753 -0.919 -0.087 0.339 0.764
1.599 2.844 4.904 8.58 9.793 11.008 12.219 12.614 14.238
15.041 15.438 16.65 17.865 18.666 20.284 21.494 23.114 24.732
26.351 28.38 31.224 34.071 37.328 40.588 45.075 50.381 55.278
61.408 65.491 71.216 76.527 84.297 91.659 96.973 105.562 113.741
121.922 130.103 130.514 134.197 138.697 145.242 150.15 156.695 163.649
170.193 180.01];
h=[94.05 94.05 94.05 107.486 107.486 107.486 107.486 120.921 67.179
53.743 94.05 94.05 107.486 107.486 107.486 107.486 94.05 94.05
120.921 134.357 147.793 174.664 174.664 228.407 295.585 362.764 403.071
604.607 752.399 913.628 1168.906 1357.006 1625.72 1907.869 2149.712
2378.119 2633.397 2902.111 3143.954 3385.797 3667.946
3936.66 4205.374 4447.217 4742.802 4998.081 5266.795 5508.637
5804.223 6072.937 6355.086 6664.107 6919.386 6811.9 6570.058
6355.086 6086.372 5844.53 5629.559 5374.28 5159.309 4904.031
4689.06 4393.474 4084.453 3788.868 3506.718 3184.261 2888.676
2606.526 2284.069 1988.484 1726.488 1504.798 1289.827
1074.856 859.885 752.399 618.042 577.735 470.25 416.507 389.635 322.457
282.15 241.843 228.407 201.536 241.843 255.278 255.278 255.278 228.407
228.407 214.971 188.1 174.664];
hcal=interp1(theta,h,thetacal);
plot(thetacal,hcal,'-r')
xlabel('Crank angle, deg')
ylabel('Heat trasnfercoefficent, W/m^2K')
title('Comparison of Woschni, Annand and Hohenberg heat transfer
cofficients')
end

functionhcal=htch(thetacal)
theta=[-179.991 -175.9 -170.991 -164.445 -156.672 -150.127 -
144.4 -139.081 -132.534 -125.58 -118.625 -110.034 -103.487 -
96.942 -90.804 -82.212 -74.029 -66.255 -58.07 -51.111 -45.788 -40.467 -
35.963 -32.685 -26.951 -22.034 -18.343 -14.245 -10.552 -5.636 -1.54 3.369
8.274 11.539 15.621 20.519 25.419 29.91 36.858 42.579 48.709
54.844 60.977 66.293 72.429 77.335 85.108 91.241 97.787 103.923
9

110.878 118.65 126.013 132.967 139.921 145.647 153.009 160.782 164.464


169.373 174.691 179.599];
h=[147.793 147.793 161.228 161.228 161.228 161.228 161.228 161.228 188.1
188.1 188.1 188.1 214.971 214.971 241.843 268.714 282.15 309.021
362.764 436.66 510.557 564.299 631.478 725.528 846.449 967.37 1128.599
1249.52 1437.62 1558.541 1652.591 1652.591 1571.977 1437.62
1289.827 1101.727 953.935 792.706 685.221 577.735 470.25 443.378
389.635 349.328 335.893 295.585 295.585 241.843 241.843 241.843 241.843
228.407 221.689 214.971 201.536 174.664 161.228 161.228 161.228 161.228
161.228 134.357];
hcal=interp1(theta,h,thetacal);
plot(thetacal,hcal,'-b')
end

functionhcal=htca(thetacal)
theta=[-180.403 -176.312 -170.585 -165.266 -159.13 -153.812 -
148.494 -142.766 -138.266 -135.814 -132.541 -127.632 -
122.313 -116.586 -110.858 -105.949 -101.449 -94.493 -87.948 -
83.038 -76.492 -70.354 -62.989 -55.623 -49.484 -44.163 -37.205 -31.881 -
27.785 -23.688 -19.181 -15.49 -10.574 -8.111 -7.284 -5.636 -3.171 -0.708
2.981 7.886 11.767 15.235 18.499 21.35 25.843 30.332 35.234
39.726 44.22 49.536 55.666 60.572 67.525 73.249 79.384 86.336
94.107 99.425 104.742 110.469 116.195 121.922 128.875 135.012 141.966
148.51 156.283 164.465 170.6 176.328 179.6];
h=[107.486 107.486 107.486 107.486 107.486 107.486 107.486 107.486 107.486
67.179 80.614 80.614 80.614 80.614 94.05 94.05 94.05 120.921
107.486 134.357 134.357 161.228 188.1 228.407 282.15 322.457 389.635
483.685 564.299 671.785 785.988 940.499 1074.856 1209.213 1370.441
1558.541 1733.205 1880.998 2001.919 1934.741 1840.691
1679.463 1531.67 1330.134 1195.777 1007.678 886.756 752.399
658.349 618.042 510.557 456.814 429.942 376.2 349.328 309.021 268.714
268.714 255.278 241.843 228.407 228.407 201.536 201.536 201.536 174.664
174.664 174.664 161.228 161.228 161.228];
hcal=interp1(theta,h,thetacal);
plot(thetacal,hcal,'-k')
end
10

Actual Graph

Predicted Graph
AN INTEGRATED MODEL FOR PREDICTING ENGINE FRICTION
LOSSES IN INTERNAL COMBUSTION ENGINE

ABSTRACT:
A comprehensive computer model (EFLP) has been developed for carrying out friction
estimation and analysis of spark ignited reciprocating engines. EFLP considers five major
losses components in an automotive engine: crankshaft, piston, valvetrain, auxiliary
components and pumping losses. Hydrodynamic, mixed, and boundary layer regimes are
used to model the friction phenomena. Empirical engine friction models based on
experimental data are the main building blocks of EFLP. EFLP is constructed using MATAB
based engine code for estimating engine performance. The friction in a 4-cylinder 4-stroke
engine is inspected.

Nomenclature

Oil viscosity...
B Cylinder bore mm cSt
0
L Engine stroke... mm Reference viscosity cSt

rc Compression ratio nv Intake or Exhaust valves /cylinder


N Engine speed... rpm ri Intake valve diameter/bore ratio
nc Number of cylinders re Exhaust valve diameter/bore ratio
nb Number of bearings Pa Atmospheric pressure kPa

Mean piston speed (2LN) m/s Pi Intake pressure... kPa
Db Bearing diameter. mm Pe Exhaust pressure kPa
Lb Bearing length. mm ff Flat follower
lvmax Maximum valve lift. mm rf Roller follower

1. INTRODUCTION:

For modelling of internal combustion engines, working with any type of fuel, developing a
model for determining the friction losses is in the top of priorities. The accuracy of such a
model has a significant importance for predicting transient behaviour of the engine and the
corresponding fuel consumption. Unfortunately, the friction losses are something unable to
overcome. Therefore, the friction minimization becomes a major consideration in the design
and operation (Heywood, 1988). The lubrication of various moving components was
described as the key to the life of the engine (Bhushan, 2002).
Almost, each automobile maker has its own investigations and study group for the friction
in the engine. It is possible with these models to estimate friction (and wear) for different
engine speeds and loads. The most well-known and used model is that of Patton et al. (1989).
Their model predicts friction mean effective pressure (FMEP) for SI engines. The model
based on a combination of fundamental scaling laws and empirical results, included
predictions of rubbing losses from the crankshaft, reciprocating, and valvetrain components,
auxiliary losses from engine accessories, and pumping losses from the intake and exhaust
systems. However, the data used to develop this model date back to the mid-1980s. With this
justification, Sandoval and Heywood (2003) modified the model of Patton et al. (1989). Their
strategy was comparing the predictions of Pattons model with more recent engine friction
data, and determining the changes that needed to be made to certain terms in this friction
model.
It was found that the pumping losses over the intake and exhaust strokes, and the terms
representing the cylinder gas pressure loading effects needed to be modified, and appropriate
adjustments were made. The authors were expanded model to include lubricant viscosity as
one of the model variables, so that the effects of component temperatures on engine friction
during cold start and warm-up transients can be predicted (Patton et al., 1989). Over the last
two decades, large gains were made, mostly by empirical methods, in reducing engine
friction and improving its reliability and durability.
Further significant improvements are possible. However, these improvements will be
more difficult than in the past. It will be necessary to have a good mathematical friction and
lubrication model for robust design of mechanical components and for understanding how
friction is developed, how it is distributed among components, and how it can be reduced. In
addition, mathematical models are necessary to respond to one of the pressing needs for the
automotive industry-faster product development. Reliable predictive models are absolutely
essential for reducing development time. This paper describes a model for friction analysis of
automotive engines using MATLAB codes.
2. ENGINE FRICTION MODEL:
The model is developed from several components. Each component estimates the friction
in a main part of the engine. The five components are the main building blocks including the
crankshaft friction, reciprocating friction, valvetrain friction, accessories friction, and
pumping losses.

2.1. Crankshaft Friction


The friction in the crankshaft is sourced from three components viz. main journal bearing,
front and rear main bearing seal and losses due to the transport of oil through the bearing. It is
calculated as Eq. (A) (Sandoval and Heywood, 2003; Ferguson and Kirkpatrick, 2001;
Heywood, 1988):
The friction in the main journal bearing of the crankshaft is calculated as Eq. (1). The
FMEP of a journal bearing array, with nb bearings, was scaled linearly with engine speed,
assuming constant bearing clearance and oil viscosity (Heywood, 1988). The proportionality
constant ( ) was suggested by Patton et al. (1989) as for the 10W-30 oil grade. The oil
properties are included in the constant Cb. The presence of the term was suggested by
Sandoval and Heywood (2003) to generalize the correlation stated in Eq. (1) and it suitable
for any other type of oil. The reference viscosity for the 10W-30 oil grade is 10.8 cSt at 100
C. The typical viscosity data for several different oil grades are listed in Table 1.
3
NDb L b n b
C
b
fmep (1)
bearings 2
0 B Ln c

Where,
Cb = Proportionality constant
= 3.03 10-4 (kPa-min/rev-mm)

Table 1: Typical test data for engine oil viscosity (Mang and Dresel, 2001)
Oil grade 5W-20 5W-30 10W-30 10W-40 20W-50
cSt at 40 C 49.2 66.1 74.8 98.9 14.4
Kinematic Viscosity
cSt at 100 C 8.6 11 10.8 14.4 19.1

The front and rear main bearing seal friction is calculated as Eq. (2). The crankshaft
bearing seals operate in a boundary lubrication regime, since the seals directly contact the
crankshaft surface. As the normal force, the seal lip load and the friction force are constant
and FMEP is independent of engine speed (Ferguson and Kirkpatrick, 2001). The
proportionality constant (Cs) was suggested by Patton et al. (1989).
D
b
fmep C (2)
seals S 2
B Ln
c
Where,
Cs = 1.22 105 (kPa-mm2)

The losses due to the transport of oil through the bearing is calculated as Eq. (3). This term
accounting for turbulent dissipation, i.e. the work required to pump the fluids through flow
restrictions (Patton et al., 1989).
N 2D2 L n
10 b b b
fmep 1.35 10 (3)
TD n
c
The total crankshaft friction is,
FMEP fmep fmep fmepTD (A)
crankshaft bearings seals

2.2. Reciprocating Friction:


The reciprocating parts produce frictions are piston skirt, piston rings and connected rod.
The friction from the piston ring has two components: one resulting from the ring tension and
the other component from the gas pressure loading.
The piston skirt friction assuming hydrodynamic regime is given by Eq. (4). The viscosity
scaling was also incorporated in this term as proposed by Sandoval and Heywood, (2003). In
addition, the piston skirt length and skirt clearance were scaled directly with the bore B
(Ferguson and Kirkpatrick, 2001). Again, the proportionality constant cps was suggested by
Patton et al. (1989); and it included oil properties.

Sp
fmep skirt C ps (4)
0 B

Where,
Cps = 294 kPa-mm-s/m
The friction component in the piston rings is due to ring tension under mixed regime and
calculated by Eq. (5). The piston ring tension and surface roughness were included to take
into account decreases in piston friction that have occurred in these two areas. The friction
coefficient was assumed to be inversely proportional to the engine speed. The value of
(1+500/N) was used to make the friction coefficient decrease by a factor of 1.8 from low to
high speeds. The proportionality constant Cpr was proposed by Patton et al. (1989).
500 1
C 1
pr
fmep (5)
rings N B 2

Where,
Cpr = 4.06 104 KPa-mm2
The hydrodynamic regime journal bearing friction from the connecting rod bearings is
given by Eq. (6). This term is similar to the bearing friction term in the crankshaft friction
term.
3
NDb L b n b
C
b
fmep (6)
bearings 2 Ln
0 B
c
A correlation for the component of piston friction due to the gas pressure loading is given
by Eq. (7). This friction also affect the total reciprocating friction. This term is a correlation
suggested originally by Bishop (1964), adopted by Patton et al. (1989) in their model and
finally modified by Sandoval and Heywood (2003). It used the product of intake pressure and
a factor which included the compression ratio and mean piston speed derived from the
physics of the compression process.

P (1.33 KS )
p
fmep 6.98 i 0.088 rc 0.128rc (7)
gasload P
a 0
Where,
K = 2.3810-2 s/m
The total reciprocating friction is,
FMEP fmep fmep fmep fmep (B)
reci skirt rings bearings gasload

2.3.Valvetrain Friction:
The term of the valvetrain friction included estimates of camshaft, cam follower, and valve
actuation mechanism friction, for a variety of valvetrain configurations (Pipitone, 2009). The
model is able to predict friction for the common types of valve configurations.
The camshaft bearing friction as a hydrodynamic regime is given by Eq. (8). It is similar to
that of the main crankshaft and connecting rod bearings friction. The constant (244 kPa-mm3-
min/rev) was suggested as the proportionality constant; plus an additional value of 4.12 kPa
to account for the camshaft seals (Patton et al., 1989).

Nn b
fmep 244 4.12 (8)
bearings B2 Ln
0 c
The valvetrain friction includes the friction resulting from relative motion between the
cam lobe and the cam follower. The friction in the mixed regime for flat follower and roller
follower configurations are different. One of the term should be used and the second must be
discarded depending on the type of the follower being investigated.
A flat follower (ff) is assumed to operate in the mixed lubrication regime, and can be
scaled with a friction coefficient inversely proportional to engine speed.

500 n v
fmep C 1 (9)
ff ff N Nn
c
A roller follower (rf) operate in the rolling contact friction regime, and can be scaled with
a friction coefficient proportional to engine speed. Patton et al, (1989) recommend the
following fmep Eq. (10),
Nn
fmep v (10)
rf Ln
c
The oscillating hydrodynamic friction, predicts friction caused by relative motion between
valvetrain components whose lubrication states were either completely or partially
hydrodynamic, such as the valve lifter in the lifter bore or the valve in the valve guide. The
friction coefficient was assumed proportional to the valve lift, and inversely proportional to
the engine speed.
1.5 0.5
l vmax N n v
fmep C
oh (11)
oh BLn
0 c
The oscillating mixed lubrication in the valve stem and tip can be calculated with a friction
coefficient proportional to valve lift, and inversely proportional to engine speed.

500 l vmax n v
C 1
om
fmep (12)
om N Ln
c

The total valvetrain friction is given by Eq. (C) as,

FMEP fmep fmep fmep fmep (C)


valvetrain bearings f oh om
The constants Cff, Crf, Coh and Com in the above equqtions are dependent of the
valvetrain configuration being considered in the model. Single Overhead Camshaft (SOHC)
and Double Overhead Camshaft (DOHC) are basically the two main ways to configure the
valvetrain. In SOHC, the camshaft is situated in the cylinder head, above the valves. The
valves are opened and closed either directly with a shim between the cam lobe and the valve
stem, or via a rocker arm. SOHC engine valve configurations typically have two or three
valves per cylinder. In the DOHC configuration, two camshafts are used in each cylinder
head. This allows to, easily, implement four valves per cylinder setup. The disadvantageous
of such setup are more weight, more cost and more complexity. Beyond this main
classification, several designs are available under each configuration. Table 2 lists the most
used types of valvetrain (Heywood, 1988). Table 3 tabulates values for the constants Cff, Crf,
Coh and Com for different valvetrain configurations.

Table 2: Valve train designs (Heywood, 1988)

Type I OHC Overhead cam; direct acing/ flat or roller follower


Type II OHC Overhead cam; End pivot rocker/ flat or roller follower
Type III OHC Overhead cam; Center pivot rocker/flat or roller follower
Type IV CIB Cam-in-block; rocker arm/flat or roller follower

Table 3: Constants for valvetrain friction term (Ferguson and Kirkpatrick, 2001)

Flat Roller Oscillating Oscillating


Follower Follower Hydrodynamic Mixed
Configuration Type Cff Crf Coh
Com
(KPa- (KPa-mm- (KPa-mm-
(KPa)
mm) min/rev) min/rev)1/2
Single overhead cam
I 200 0.0076 0.5 107
(SOHC)
Double overhead cam
I 133 0.0050 0.5 10.7
(DOHC)
Single overhead cam
II 600 0.0227 0.2 42.8
(SOHC)
Single overhead cam
III 400 0.0151 0.5 21.4
(SOHC)
Cam in block (CIB) IV 400 0.0151 0.5 32.1
2.4. Accessory Friction:
The component of the accessory (or auxiliary) friction term is an empirical match to the sum
of oil pump, water pump, and non-charging alternator friction. All of the auxiliary component
friction torques was assumed to be proportional to the engine displacement. Sandoval and
Heywood (2003) suggest an accessory mean effective pressure FMEPaux of the form:

fmep 8.25 1.86 10 3 N 7.45 10 7 N 2 (D)


aux

2.5. Pumping Losses:


The pumping mean effective pressure is the sum of the pressure drop across flow
restrictions during the intake and exhaust strokes. It is the measure of work required to move
the fuel-air mixture into and out of an engine. The flow restrictions are located in the intake
system, the intake valves, the exhaust valves, and the exhaust system. The intake system
elements include the air filter, intake manifold, and throttle valve. The exhaust system flow
restrictions include the exhaust manifold, catalytic converter, muffler, and tail pipe.
The intake manifold vacuum, calculated as the difference between atmospheric and intake
pressure.
P P P (13)
im a i
The intake port and valve pressure drop is calculated as,

2 S 2
P p
P 4.12 10 3 i (14)
iv P n2 r2
a v i

The exhaust system pressure drop is calculated as,
2
P
P 0.178 i S (15)
es P p
a
The exhaust port and valve pressure drop is calculated as,

2 S 2
P p
P 4.12 10 3 i (16)
ev P n 2r 2
a v e

Where,
The total pumping mean effective pressure is,
FMEP P P P P (E)
pumping im iv es ev
3. Engine Friction Losses of 4-cylinder 4-stroke engine:
Reducing the friction losses and wear of the parts subjected to relative motion is a main
aim in the engine design. Engine speed is the major operation parameter that controls the
friction losses. In this section the attention is, therefore, paid to explore the trends of the
different components of friction losses with engine speed, according to the predictions of
EFLP. The results are presented here for an engine with the key parameters listed in Table 4.
The friction behaviour of the different engine components are presented in the following
section. The contribution of each individual part is highlighted to inspect the weight of each
part.
Table 4 lists the specifications of 4-cylinder 4-stroke engine and Engine Friction Losses
are constructed using MATAB based engine code for estimating engine performance.
Table 4. Engine parameters of the case study (Heywood, 1988)
Parameter 4 - Cylinder Engine
Bore, B (mm) 86
Stroke, L (mm) 86
Number of cylinders, nc 4
Compression ratio, rc 9
Atmospheric pressure, Pa (kPa) 101
Intake pressure, Pi (kPa) 101
Exhaust pressure, Pe (kPa) 103
Number of connecting rod bearings, nb 4
Connecting rod bearing diameter, Db (mm) 48
Connecting rod bearing length, Lb (mm) 42
Intake valves/cylinder, nv 2
Exhaust valves/cylinder, nv 2
Intake valve diameter (mm) 35
Exhaust valve diameter (mm) 31
Maximum valve lift, lvmax (mm) 11
Number of crankshaft bearings, nb 5
Crankshaft bearing diameter, Db (mm) 56
Crankshaft bearing length, Lb (mm) 21
Number of camshafts 1
Valvetrain type SOHC type I
Figure 1 represents the trends of the crankshaft friction. Clearly, the friction loss of the
crankshaft bearing seal does not depend on the engine speed and remains constant for all the
investigated speed range. In 4-cylinder engine, 70% of the total crankshaft friction was
accounted for by the hydrodynamic and turbulent dissipation term.

Figure 1: Crankshaft Friction behaviour with engine speed

In Figure 2, piston friction and its elements are pictured. The skirt and connecting rod
bearing FMEP increase linearly with engine speed, while the piston rings and gas pressure
FMEP decrease with engine speed. At higher speeds the majority of the friction is from the
piston skirt. This is because piston skirt is designed to meet the side thrust forces originating
from the rotation of the connecting rod; which augmented upon increasing piston speed. The
connecting rod bearing and gas pressure FMEP was more.

Figure 3 shows the friction losses in the valvetrain. It is recognizable that valvetrain is a
main source of friction losses in the engine over the entire engine speed range. This is due the
high loads carried by the valvetrain over the entire speed range. More specifically, at lower
speeds, valvetrain is under loads primarily due to spring forces, while at higher speeds the
inertia forces of the component masses dominate. Also, it can be seen from the shape of the
FEMPvt versus speed curve that the cam follower interface contributes the largest friction loss
due to the very high loads and small contact areas. The other characterized feature seen here
is the higher losses at lower speeds which is contradictory to the other component behaviour.

Figure 2: Piston friction behaviour with engine speed

Figure 3: Valvetrain friction behaviour with engine speed


Figure 4 describes the constituents of the pumping losses friction. The pumping mean
effective pressure is the sum of the pressure drops across flow restrictions during the intake
and exhaust strokes. It is a measure of the work required to move the fuel-air mixture into and
out of an engine (Ferguson and Kirkpatrick, 2001). The major part for the tested condition is
the losses in the intake system which depends on the manifold and ambient pressures only.
The fmep terms were increases gradually with speed.

Figure 4: Pumping losses behaviour with engine speed

Figure 5 presents the total engine FMEP into its components and shows the weight of each
component. Evidently, all the losses components increase with engine speed except that of
the valvetrain. In addition, the main contributor is the piston friction.
Figure 5: Engine friction components behavior with engine speed

CONCLUSIONS:
This paper gives an overview of EFLP, a comprehensive computer analysis model for
friction analysis of automotive engines. Details on the mathematical formulation of the
friction phenomenon were described. EFLP provides a unique tool to examine impact of
engine design changes on engine friction. Because of the difficulty in accurately measuring
friction in a running engine, a comprehensive mathematical tool, such as the one described in
this paper, is invaluable in designing inherently low-friction engines and in minimizing
friction for existing designs.

REFERENCES:
Mohammed Kamil, M.M. Rahman and Rosli A. Bakar, AN INTEGRATED MODEL
FOR PREDICTING ENGINE FRICTION LOSSES IN INTERNAL
COMBUSTION, 1-3 July 2013
Ferguson, C.R. and Kirkpatrick, A.T. 2001. Internal combustion engine: applied
thermodynamic. Second edition. New York: John Wiley and Sons
MATLAB PROGRAMS:
1. Crankshaft friction:
cb = 3.03E-4;cs = 1.22E5;mu = 10.8;mu0 = 10.8;Db = 56;Lb = 21;nb = 5;
nc = 4;B = 86;L = 86;

N = 1000:10:6000;
% Main Journal Bearing
MJB = cb * sqrt(mu/mu0) * ( N * Db^3 * Lb * nb /(B^2 * L * nc));
% Front & Rare Main Bearing Seal
MBS = cs * Db / (B^2 * L * nc);
% Transport of Oil
TD = 1.35E-10 * (N.^2 * Db^2 * nb / (nc) ) ;

% CR1=41.37*0.14*(B/L)*(N/1000);

plot(N, MBS, '-r','linewidth',2) hold on


plot(N, MBS+MJB, '-m','linewidth',2) hold on
plot(N, TD+MBS+MJB, '-b','linewidth',2)

text(2500,1.2,'\fontsize{12}Main Crankshaft Bearing Seal')


text(2500,6,'\fontsize{12}Main Crankshaft Bearing Hydrodynamic')
text(4100,18,'\fontsize{12}Turbulent Dissipation')
text(2000,25,'\fontsize{12}Total Crankshaft Friction')

xlabel('Engine Speed N (RPM)','FontWeight','bold')


ylabel('Friction Mean Effective Pressure (kPa)','FontWeight','bold')
title('4-Cylinder 4-Stroke Engine','FontWeight','bold')

2. Reciprocating friction
mu= 10.8;mu0 = 10.8;Db = 48;Lb = 42;nb = 4;nc = 4;B = 86;L = 86;
N = 1000:10:6000;Sp = 2*L.*N/60000;
Cps = 294;Cpr = 4.06E4;Pi = 66;Pa = 101;rc = 9;K = 2.38E-2;

% Piston Skirt Friction


SK = Cps * sqrt(mu/mu0) * (Sp/B);
% Piston Rings Friction
PR = Cpr * (1+(1000./N)) * (1/B^2);
% Connecting Rod Beaarings friction
CR = 3.03E-4 * sqrt(mu/mu0) * ( N.* Db^3 *Lb * nb / (B^2*L*nc));
% Gas Pressure
GL = 6.89* (Pi/Pa)* ( (0.088* sqrt(mu/mu0)* rc) + (0.182* (rc .^(1.33 -
(K .* Sp))))) ;
% Total Piston Friction
% FMEP_Rec= SK+PR+CR+GL;

plot(N,SK,'r','linewidth',2) hold on
plot(N,PR+SK,'-b','linewidth',2) hold on
plot(N,CR+PR+SK,'-m','linewidth',2) hold on
plot(N,GL+CR+PR+SK,'k','linewidth',2)

text(4000,20,'\fontsize{12}Piston Skirt')
text(1600,23,'\fontsize{12}Piston')
text(1450,20,'\fontsize{12}Rings')
text(5100,65,'\fontsize{12}Bearings')
text(1300,38,'\fontsize{12}Gas')
text(1250,34,'\fontsize{12}Pressure')
text(2500,80,'\fontsize{12}Total Piston Friction')

title('4-Cylinder 4-Stroke Engine','FontWeight','bold')


xlabel('Engine Speed N (RPM)','FontWeight','bold')
ylabel('Friction Mean Effective Pressure (kPa)','FontWeight','bold')

3. Valvetrain friction:

mu = 10.8; mu0 = 10.8;Db = 56;Lb = 21;nb = 5;nc = 4;B = 86; L = 86;


lv=11;nv=2;

N= 1000:10:6000;Sp=2*L.*N/60000;
% Constants for SOHC
Cff=200;Crf=0.0076;Coh=0.5;Com=10.7;

% Camshaft Bearing Friction


CB = (244* sqrt(mu/mu0)*(( N.*nb )/(B^2*L*nc))) + 4.12;
% Contact friction for Flat Follower
FF = 8 * Cff * (1+(1000./N)) * (nv/(L*nc));
% Contact friction for Roller Follower
% RF = Crf * (N.*nv/(L*nc));
% Oscillating Hydrodynamic Friction
OH = 8*Coh * sqrt(mu/mu0)* (lv^1.5 * N.^0.5 * nv / (B* L* nc));
% Oscillating Mixed Friction
OM = 8*Com * (1+(1000./N)) * ((lv * nv) / (L * nc));
% Total Valvetrain Friction
% VT=CB+FF+OH+OM; or VT=CB+RF+OH+OM;

plot(N,CB,'-r','linewidth',2) hold on
plot(N,FF+CB,'-m','linewidth',2) hold on
plot(N,OH+FF+CB,'--','linewidth',2) hold on
plot(N,OM+OH+FF+CB,'-k','linewidth',2)

text(3200,3,'\fontsize{12}Crankshaft Bearings')
text(3200,12,'\fontsize{12}Cam Follower')
text(5000,18.1,'Oscillating Hydrodynamic')
text(3200,22,'\fontsize{12}Oscillating Mixed Friction')
text(3200,31,'\fontsize{12}Total Valvetrain Friction')

xlabel('Engine Speed N (RPM)','FontWeight','bold')


ylabel('Friction Mean Effective Pressure (kPa)','FontWeight','bold')
title('4-Cylinder 4-Stroke Engine','FontWeight','bold')

4. Pumping Losses:
L=86;B=86;Di=35;De=31;ri=Di/B;re=De/B;nv=2;Pi=66;Pa=101;
N= 1000:10:6000;Sp=2*L.*N/60000; % Mean Engine Speed

% Intake Manifold Vacuum


V=Pa-Pi;
% Intake port and valve Pressure Drop
IV=4.12*10^-3 * ((Pi/Pa)^2) * ((Sp.^2) / (nv^2*ri^4));
% Exhaust system Pressure Drop
ES= 0.178* ((Pi/Pa)*Sp).^2;
% Exhaust port and valve Pressure Drop
EV=4.12*10^-3*(Pi/Pa)^2*((Sp.^2) / (nv^2*re^4));
% Total Pumping Losses
% PT=V+IV+ES+EV;

plot(N,V,'-r','linewidth',2) hold on
plot(N,IV+V,'-b','linewidth',2) hold on
plot(N,ES+IV+V,'-m','linewidth',2) hold on
plot(N,EV+ES+IV+V,'-k','linewidth',2)

text(2800,19,'\fontsize{12}Intake Manifold Vacuum')


text(4200,36.3,'\fontsize{12}Intake port & valve pressure drop')
text(3800,43,'\fontsize{12}Exhaust system Pressure Drop')
text(5500,63,'\fontsize{12}Exhaust port')
text(5300,60,'\fontsize{12}& valve')
text(3000,70,'\fontsize{12}Total Pumping Losses')

axis([1000 6000 0 80])


xlabel('Engine Speed N (RPM)','FontWeight','bold')
ylabel('Friction Mean Effective Pressure (kPa)','FontWeight','bold')
title('4-Cylinder 4-Stroke Engine','FontWeight','bold')

5. Total Friction of engine:

B =86;L = 86;nc = 4;nv=2;lv=11;m = 10.8;m0 = 10.8;Pi=66;Pa=101;

N= 1000:10:6000; Sp=2*L.*N/60000;

% crankshaft friction
cb = 3.03*10^-4; cs = 1.22*10^5;Db = 56;Lb = 21;nb = 5;

MJB = cb * sqrt(m/m0) * ( N* Db^3 *Lb*nb /(B^2*L*nc));


MBS = cs * Db / (B^2 * L * nc);
TD = (1.35*10^-10) * (N.^2 * Db^2 * nb / nc) ;

CST = MJB + MBS + TD ;


plot(N,CST,'linewidth',2) hold on

% Valvetrain friction
Cps=294; Cpr=4.06*10^4; Cff=200; Crf=0.0076;Coh=0.5;Com=10.7;nb3 = 5;

CB=(244* sqrt(m/m0)*( N.*nb3 /(B^2*L*nc))) +4.12;


FF=Cff* (1+(500./N)) * (nv/L*nc);
OH=Coh * sqrt(m/m0)* (lv^1.5 * N.^0.5 * nv / (B* L* nc));
OM=Com * (1+(500./N)) * (lv * nv / (L * nc));

VT=CB+FF+OH+OM;

% Reciprocating friction
Cps=294;Cpr=4.06*10^4;Lb = 42;Db2 = 48;nb2 = 4;rc=9;K=2.38*10^-2;

PF=Cps * sqrt(m/m0) * (Sp/B);


PR= Cpr * (1+(1000./N)) * (1/B^2);
CR= 3.03*10^-4 * sqrt(m/m0) * ( N.* Db2^3* Lb * nb2 /(B^2*L*nc));
GL= 6.89* (Pi/Pa)* ( (0.088*sqrt(m/m0)*rc) + (0.182* (rc .^(1.33-(K .*
Sp)))) );

RT= PF+PR+CR;
plot(N,CST+VT+RT,'--','linewidth',2) hold on
plot(N,RT+CST,'-r','linewidth',2) hold on

% Accessory Friction
AT= 8.32+ (1.86*10^-3*N) + (7.45*10^-7*N.^2);

plot(N,AT+VT+RT+CST,'-m','linewidth',2) hold on

% Pumping Losses
Di=35; De=31; ri=Di/B;re=De/B;Pi=66;Pa=101;

V=Pa-Pi;
IV=4.12E-3* (Pi/Pa)^2* (Sp.^2 / (nv^2*ri^4));
ES= 0.178* ((Pi/Pa)*Sp).^2;
EV=4.12E-3*(Pi/Pa)^2*(Sp.^2 / (nv^2*re^4));

PT=V+IV+ES+EV;
plot(N,PT+AT+VT+RT+CST,'-k','linewidth',2)

text(2500,250,'\fontsize{12}Total Engine Friction','FontWeight','bold')


text(4700,197,'\fontsize{12}Pumping Losses')
text(4820,140,'\fontsize{12}Auxilliary')
text(1250,52,'\fontsize{12}Reciprocating')
text(1350,24,'\fontsize{12}Valvetrain Friction')
text(4250,14,'\fontsize{12}Crankshaft Friction')

title('4-Cylinder 4-Stroke Engine','FontWeight','bold')


xlabel('Engine Speed N (RPM)','FontWeight','bold')
ylabel('Friction Mean Effective Pressure (kPa)','FontWeight','bold')
1

DYNAMIC FORCE ANALYSIS

Dynamic analysis of Slider Crank Mechanism


A typical crank -slider mechanism as used in an IC Engine is shown in Fig. 1. It essentially consists of
four different parts viz., frame (i.e., cylinder), crank, connecting rod and reciprocating piston. The frame
is supposedly stationary; crank is undergoing purely rotary motion while the piston undergoes to-and-fro
rectilinear motion. The connecting rod undergoes complex motion its one end is connected to the crank
(undergoing pure rotation) and the other end is connected to the piston (undergoing pure translation).

We know that the inertia forces are given by mass times acceleration and we shall now estimate the
inertia forces (shaking forces and moments) due to the moving parts on the frame (cylinder block).

Fig. Schematic of Slider-Crank Mechanism


Piston
From the geometrical construction shown, we have
2

x r[(n 1) (n cos cos )]


x r (1 cos ) (n n 2 sin 2 )
x = displacement of piston from inner dead center

If the connecting rod is very large n2 is very large, and hence n2 sin2 will approach n.

The equation converts to


x = r (1cos)
This is the expression for a SHM. Thus the piston executes SHM when connecting rod is large.

Velocity of Piston
dx dx d
v .
dt d dt
sin 2
v r sin
2n 2 sin 2

if n2 is large compared to sin2 then


sin 2
v r sin
2n
sin 2
If can be neglected (when is quite less) then
2n
v r sin

Acceleration of Piston
dv dv d
a .
dt dt dt

d sin 2 2 cos 2
a
d r sin 2n r cos n

3

Fig. Variation of Primary and Secondary Inertia Forces

Crank

Crank undergoes pure rotary motion and let us assume that it is rotating at a constant speed . Let G be
the center of mass of the crank as shown in Fig 3.3 and so we can write:

Fig. Typical Crank


4

Centripetal acceleration = (OG) 2

The inertial force (radial) due to crank rotation is given by:


Inertia Force= mc(OG)2
where mc is the mass of the crank

This inertia force can be balanced out by the techniques learnt in the previous module and we could
make sure that G coincides with O. Then the inertia force would be reduced ideally to zero. If need be,
we could always use the above formula to estimate the inertia forces due to crank.

For reasons that will become clear when we discuss the dynamics of connecting rod, it is common to
assume that the entire mass of the crank is actually concentrated at the pin A. Thus we can write,

Effective crank mass at A = (OG)/(OA)

The negative sign indicates that the sense of angular acceleration of the rod is such that it tends to reduce the angle .

Net or Effective force on the Piston:-

A1 = area of the cover end A2 = area of the piston end

P1 = pressure of the cover end P2 = pressure of the piston end

m = mass of the reciprocating parts


Fp P1 A1 P2 A2
cos 2
Fb ma mr 2 cos
n

Fp= Force On Piston


Fb= Inertia Force
5

Connecting Rod

One end of the connecting rod is circling while the other end is reciprocating and any point in between
moves in an ellipse. It is conceivable that we derive a general expression for the acceleration of any point
on the connecting rod and hence estimate the inertia forces due to an elemental mass associated with that
point. Integration over the whole length of the connecting rod yields the total inertia force due to the
entire connecting rod. Instead we try to arrive at a simplified model of the connecting rod by replacing it
with a dynamically equivalent link

Fig. Dynamically Equivalent Link to Connecting Rod


6

m mA mB
mA ( AG ) mB (GB)
J mA ( AG ) 2 mB (GB) 2 J AB
For convenience we would like the equivalent link lumped masses to be located at the big and small end
of the original connecting rod and if its center of mass (G) location is to remain same as that of original

rod, distances AG and GB are fixed. Given the mass m and mass moment of inertia of the original

connecting rod, the problem of finding dynamically equivalent link is to determine , and .

An approximate equivalent link can be found by simply ignoring and treating just the two lumped

masses and connected by a mass-less link as the equivalent of original connecting rod. In such a
case we take
mA m(GB)/L
mB m(AG)/L

Thus the connecting rod is replaced by two masses at either end (pin joints A and B) of the original rod.

rotates along with the crank while purely translates along with the piston. It is for this reason
that we proposed use of crank's effective rotating mass located at pin A, which can now be simply added
up to part of connecting rod's mass.

On the shop floor , can be immediately determined by mounting the existing connecting rod on

two weighing balances located at A and B respectively. The readings of the two balance give and

directly.

How To Run The Program:


1) Load the text file containing data for p-theta diagram.
2) Save the parameters of text file into new variables.
3) Make the code and run the program.
7

MATLAB Program:

% This program shows the relation between gas and inertia forces
% graphically.
clc
clear all
load pressure_values.txt; %loads the file containing data for p-theta diagram
CA= pressure_values(:,1); %Saves the values of crank angle
P=pressure_values(:,2); %Saves the values of pressure
mcp= input('Enter mass of crankpin(in kg) \n');
mp= input('Enter mass of piston(in kg) \n');
l=input('Enter length of connecting rod(in mm) \n');
r=input('Enter length of crank rod(in mm) \n');
k=input('Enter no of cylinders \n');
a=l/r;
mcr.rot =input('Enter rotating mass of connecting rod(in kg) \n');
mcr.osc =input('Enter oscillating mass of connecting rod(in kg) \n');
d =input('Enter the diameter of piston(in m) \n');
rpm =input('Enter engine rpm \n');
rotm= mcr.rot + mcp;
oscm= mcr.osc + mp;
w=2*pi*rpm/60;
F.m.rot = rotm*r*w*w; %Calculates rotating Force
theta(1:720)=zeros;
Gas_force(1:720)= zeros;
Force_piston(1:720)=zeros;
xlim ([0 720]);
for i= 1:1:720
theta(i)=i;
8

F.m.osc(i) = (oscm*r*w*w)*((cosd(i)) + cosd(2*(i))/k); %Calculates oscillating inertia force


Gas_force(i)=((pi/4)*d*d*P(i)); %Calculates force due to gases on piston
Force_piston(i) = F.m.osc(i) + Gas_force(i); %Calculates total force on piston
end
figure(1)
plot(CA,P,'LineWidth',1.5) %plots the graph
grid on;
xlabel('Crank angle theta(in degrees)');
ylabel('Pressure(in N/m2)');
title('Combustion Pressure of a 4-stroke reciprocating engine');
legend('Pressure','Location','northeast');
figure(2)
plot(theta,Gas_force,'g','LineWidth',1.5) %plots the graph
grid on;
xlabel('Crank angle theta(in degrees)');
ylabel('Gas Force(in N)');
title('Force due to gases on piston in a 4-stroke reciprocating engine');
legend('Gas Force','Location','northeast');
figure(3)
plot(theta,F.m.osc,'r','LineWidth',1.5) %plots the graph
grid on;
xlabel('Crank angle theta(in degrees)');
ylabel('Oscillating inertia Force(in N)');
title('Oscillating inertia Force of a 4-stroke reciprocating engine');
legend('Oscillating inertia Force','Location','northeast');
figure(4)
plot(theta,F.m.rot,'k','LineWidth',1.5) %plots the graph
grid on;
xlabel('Crank angle theta(in degrees)');
ylabel('Rotating inertia Force(in N)');
title('Rotating inertia Force of a 4-stroke reciprocating engine');
9

legend('Rotating inertia Force','Location','northeast');


figure(5)
plot(theta,Force_piston,'m','LineWidth',1.5) %plots the graph
grid on;
xlabel('Crank angle theta(in degrees)');
ylabel('Force on piston(in N)');
title('Total Force on piston of a 4-stroke reciprocating engine');
legend('Force on piston','Location','northeast');

Sample Input:
Enter mass of crankpin (in kg) =1.33
Enter mass of piston (in kg) =0.8
Enter length of connecting rod (in mm) =0.45
Enter length of crank rod (in mm) =0.15
Enter no of cylinders =3
Enter rotating mass of connecting rod (in kg) =3.567
Enter oscillating mass of connecting rod (in kg) =3.33
Enter the diameter of piston (in m) =0.115
Enter engine rpm =1000
10

Output:

Fig 1. Graph showing relation between in-cylinder pressure and crank angle
11

Fig 2. Graph showing relation between in-cylinder Force due to gases and crank angle

Fig.3.Graph showing relation between Oscillating Inertia Force and crank angle
12

Fig4.. Graph showing relation between Rotating Inertia Forces and crank angle

Fig 5. Graph showing relation between total Force on Piston and crank angle
1

Instantaneous Friction Torque Model On Direct Injection Four Stroke Diesel


Engine: A Critical Review

Figure 1. shows the piston-crank mechanism with approximate kinetically equivalent


point masses replacing the connecting rod. The model includes piston pin offset for the piston
acceleration. Important geometrical parameters are the crankshaft angular position 1, measured
from where the crankshaft and the connecting rod are in a straight line; the angle of connecting
rod ; the crank radius, r, which is equal to half of the stroke; the connecting rod length, L; the
piston pin offset, and the connecting rod angle when piston is at top dead centre, . The piston
pin is slightly offset in order to reduce engine noise and wear during the change of the direction
of normal force on the piston at the end of compression.

Fig.1 force and acceleration of the piston crank mechanism

The relationship between the indicated gas pressure, pI, and the indicated torque, TI, is
deterministic and is a function of engine geometry. This relationship is expressed as
2

TI ( pI patm ) AprG(1 ) (1)


Where,

sin(1 ) 1 (1 )
G(1 ) sin 1 cos 1
cos ( 1 ) (2)
( 1) is geometric function is given by,

r sin(1 )
2

(1 ) 1
L (3)
Where,
r sin(1 )
sin 1 and sin 1 (4)
r L L
FRICTION MODELLING:
The main aim of this report is to model the friction torque given by the expression
6=1 .The friction torque plays a dominant roles within the engine dynamic equation, and
accurate modelling of this component is very important.
The experimental results used for the validation of the model were obtained from engines
using SAE 30 oil. The variation of temperature and hence oil viscosity is an important factor in
the engine friction calculations. The viscosity of oil decreases with increasing temperature. This
trend is typical for Newtonian regimes for oils, where the viscosity is independent of the shear
rate. Furthermore, the viscosity of oils increases with pressure. At certain times during
hydrodynamic lubrication, the loads increase the oil pressure, which in turn increases the
viscosity.

ENGINE FRICTION MODEL


Engine friction can be modelled to give either mean or instantaneous values. Studies
providing mean frictional losses over each cycle include, for example. There are two important
instantaneous engine friction evaluation. The first paper determines the total instantaneous
frictional torque by measuring the engine speed and the indicated pressure, but does not include
the effect of temperature variation on oil viscosity, and the ring assembly friction analysis is not
detailed enough to describe the friction behaviour during lubrication modes. The second study
presents a frictional model with three components, ring assembly, loaded bearing and auxiliaries,
3

but neglects valve train, piston skirt and camshaft losses leading to a simplified engine friction
model. The aim of this work is to build a comprehensive friction model for Diesel engines.
Theoretical studies are employed to derive engine friction components in order to predict the
engine behaviour during all of the engine working modes, from start until high engine speed.

1) PISTON ASSEMBLY FRICTION:


The piston-ring assembly may be responsible for 50-75% of the entire engine friction
a) RING ASSEMBLY FRICTION TORQUE:

The piston assembly friction is dominated by the ring friction components. It has been
observed that the ring assembly lubrication mode is hydrodynamic.except near the top and
bottom dead centre positions where the oil film breaks down. The ring assembly friction torques
can be found by assuming that the friction force is equal to the product of the normal load
between the ring assembly and the liner, and the friction coefficient. Thus the ring assembly
friction torque.


2

pi p atm d MG1 (1 ) 1
N Ei g c N
Tf 1 r G (1 ) d r Bi ai pi p atm d r Bi 4
d
3

i 1
r
i 1

7.07 d r B 1
i
(5)
Where,

L 1 ( r sin(1 ) ) 2
L
G3 (1 ) (6)
r sin(1 )


The coefficient of friction for hydrodynamic lubrication, , is directly proportional to the piston
speed and oil viscosity, and inversely proportional to the ring load.
4

c1 (c1 z ) sin 1
(7)
z

where c1 is the mixed friction coefficient for rings and z is the hydrodynamic friction coefficient
given by:




1 r G (1 )
z
Lr

(8)
and Lr, is the load per unit length given by,



N Ei g
Lr 3
PI Patm Bi
i 1 dr
7.07 d r
B 1

i (9)
b) PISTON SKIRT FRICTION TORQUE

In reciprocating internal combustion engines, the piston skirt profile is designed to prevent
local contact due to thermal expansion and to reduce friction losses and piston slap, by ensuring
that a proper lubricating oil film thickness is maintained for maximum wedge effect and an
appropriate piston pin offset. The piston skirt friction force is derived by applying Newton's law
for viscous friction and can be expressed as:

.
1 rG(1 )
Ff 2 dLs
Oc
(10)

and the piston skirt friction torque is given as:


1 rG(1 )
Tf 2
Oc dLs rG(1 )
(11)
5

2) VALVE TRAIN FRICTION TORQUE


The valve train carries high loads over the entire speed range of the engine. Loads
acting on the valve train at lower speeds are primarily due to the spring forces, while at
higher speeds the inertia forces of the component masses dominate. Valve train friction
accounts for the friction in cam/follower, rocker arm/pivot shaft, camshaft bearings,
valve/valve guides, and seals. To simplify the analysis, the valve train is considered as a
system represented by the cam/follower and is assumed to operate in the boundary
lubrication mode. The valve train friction torque is formulated to depend on the load and
the engine speed, and is given by:
d i 4 Gv
Vl N v 1 c3 1 r G(1 )

Tf 5
3
8d o N t (12)
Where, c3 is a speed coefficient.

3) AUXILIARIES AND GEAR TORQUE LOSSES

This component models the friction due to the unloaded journal bearings, water pump,
fuel pump, oil pump, generator and gears. A hydrodynamic lubrication mode is assumed, and the
friction coefficient is calculated. If the terms which represent the normal load, oil film pressure
and the effective radius and length of the auxiliaries shafts are assumed to be unity, then the
friction torque for the auxiliaries is formulated as:


Tf 6 1

There are three empirical coefficients whose values need to be assigned. Coefficient cl is
the empirical coefficient of the friction coefficient for ring assembly with mixed lubrication, its
variation can be approximated by a straight line from the Stribeck diagram without significant
error. Coefficient c2 is the empirical coefficient of friction coefficient for the friction of the
bearing during mixed lubrication, and depends on the properties of the pair of rubbing materials.
The lubrication in the valve train at lower speeds is in the boundary mode and at high speeds in
the mixed mode, so the friction is slightly decreased at high speed.
6

Mixed Friction Coefficient For rings c1 0.11


Mixed friction coefficient for bearings c2 0.002
Speed effect in valve train friction c3 0.00127(s/rad)

Table 1.Empirical coefficients

Below graphs show the instantaneous friction torque obtained from the simulation. Note
that for this plot and all other plots in this , one crank shaft cycle represents 4 radians, i. e., two
revolutions. The lubrication mode of the ring assemblies is considered to be mixed lubrication
near top dead centre. The ring assembly friction torques reach the maximum value after top dead
centre where the indicated torque is maximum and its value is approximately equal to twice the
ring friction torque at the beginning of the combustion. The lubrication mode of the skirt is in
general hydrodynamic and its friction torque is high in the mid-region of the stroke where the
piston speed is maximum. The bearing lubrication mode is hydrodynamic except at the points
where the indicated pressure is maximum, when there is mixed lubrication and the friction torque
is increased substantially; its value is approximately five times the hydrodynamic lubrication
friction torque value. The valve train friction torque increases during the intake and exhaust
strokes due to the spring stiffness forces being increased.
7

SIMULATION RESULTS

REFERANCES:
1) Nonlinear Modelling and Estimation for Diesel Engines By: Yahya H.Zweiri Kings
College LONDON Founded 1829 Department of Mechanical Engineering, King's
College London.
2) Internal combustion engine fundamentals by J.B.Heywood.
1

FLOW PAST THROTTLE PLATE

Except at or close to wide-open throttle, the throttle provides the minimum flow area in the entire
intake system. Under typical road-load conditions, more than 90 percent of the total pressure loss
occurs across the throttle plate. The minimum-to-maximum flow area ratio is large typically of
order 100. Throttle plate geometry and parameters are illustrated in following figure 1.

Figure 1:Throttle plate geometry

A throttle plate of conventional design such as Fig.1 creates a three-dimensional flow field. At
part throttle operating conditions the throttle plate angle is in the 20 to 45 range and jets issue
from the "crescent moon shaped open areas on either side of the throttle plate. The jets are
primarily two dimensional immediately below the throttle plate.

1
2

Figure 2:Photographs of flow in two-dimensional hydraulic analog of carburetor venturi, throttle


plate, and manifold plenum floor at different throttle plate angles.
Figure 2 shows photographs taken of a two-dimensional hydraulic analog of a typical carburetor
venturi and throttle plate in steady flow at different throttle angles. The path lines traced by the
particles in the flow indicate the relative magnitude of the flow velocity. The flow accelerates
through the carburetor venturi (separation occurs at the corners of the entrance section); it then
divides on either side of the throttle plate. There is a stagnation point on the upstream side of the
throttle plate. The wake of the throttle plate contains two vortices which rotate in opposite
directions. The jets on either side of the wake (at part throttle) are at or near sonic velocity. There
is little or no mixing between the two jets. Thus, if maldistribution of the fuel-air mixture occurs
above the throttle plate, it is not corrected

In analyzing the flow through the throttle plate, the following factors should be considered:
1. The throttle plate shaft is usually of sufficient size to affect the throttle open area.
2. To prevent binding in the throttle bore, the throttle plate is usually completely closed at
some nonzero angle (5, 10, or 15).
3. The discharge coefficient of the throttle plate is less than that of a smooth

2
3

4. Converging diverging nozzle, and varies with throttle angle, pressure ratio, and throttle
plate Reynolds number.
5. Due to the manufacturing tolerances involved, there is usually some minimum leakage
area even when the throttle plate is closed against the throttle bore. This leakage area can
be significant at small throttle openings.
6. The measured pressure drop across the throttle depends (+ 10 percent) on the
circumferential location of the downstream pressure tap.
7. The pressure loss across the throttle plate under the actual flow conditions (which are
unsteady even when the engine speed and load are constant) may be less than under
steady flow conditions

Consider the flow of an ideal gas with constant specific heats through the duct. For the ideal
flow, the stagnation temperature and pressure T0 and P0 are related to the conditions at other
locations in the duct by the steady flow energy equation
V2
T0 T
2C p
(1)
Where T is temperature and V is velocity of air at intake manifold side

And the isentropic relation


1
T Pt

T0 P0 (2)

By introducing the Mach number M = V/Vs, where Vs is the sound speed



RT , the

following equations are obtained:

T0 1 2
1 M
T 2 (3)

3
4


P0 1 2 1
1 M
Pt 2 (4)

The mass flow rate mideal is


mideal AthV (5)

With the ideal gas law and the above relations for Pt and T, this can be rearranged as

1
mideal RT0 1 2 2 1
M 1 M
Ath P0 2
(or)

1 1
mideal RT0 P 1 2 2 1
t 1 M
Ath P0 P0 2 (6)

For given values of Po and To, the maximum mass flow occurs when the velocity at the
minimum area or throat equals the velocity of sound. This condition is called choked or
critical flow. When the flow is choked the pressure at the throat Pt is related to the stagnation
pressure P0 as follows
Critical Value

2 1

1 (7)
For a real gas flow, the discharge coefficient is introduced. The throttle plate discharge
coefficient (which varies with Ath and minimum leakage area, must be determined
experimentally. The mass flow rate through the throttle valve can be calculated from standard
orifice equations for compressible fluid flow. For pressure ratios across the throttle more than the
critical value (Pt/P0 =0.528; for = .4 , the mass flow rate is given by

4
5


Mth=
(8)

+

Where K=

[

] (for
> Critical value)

Where Ath is the throttle open area, Po and To are the upstream Pressure and temperature, pt is
the pressure downstream of the throttle plate (assumed equal to the pressure at the minimum area
(i.e. no pressure recovery occurs), and Cd is the discharge coefficient (determined
experimentally). For pressure ratios less than the critical ratio, when the flow at the throttle plate
is choked


Mth=

+

Where K= [ +
+
] (for
< Critical value) (9)

The throttle plate open area Ath as a function of angle for the geometry in Fig. 1 is given by

4 cos cos cos


/
= cos + cos cos cos cos sin ( )
cos

/
+ sin (10)

Where a = d/D, d is the throttle shaft diameter, D is the throttle bore diameter and 0 is the
throttle plate angle when tightly closed against the throttle bore When = cos a cos the
throttle open area reaches its maximum value ( D2/4 dD).

The relation between air flow rates, throttle angle, intake manifold pressure, and engine
speed for a two-barrel carburetor and a 4.7-dm3 (288-in3) displacement eight-cylinder
production engine is shown in Figure3. While the lines are from a quasi-steady computer
simulation, the agreement with data is excellent. The figure shows that for an intake manifold
pressure below the critical Value (0.528 x Patm = 53.5 kN/mZ = 40.1 cm of Hg) the air flow rate
5
6

at a given throttle position is independent of manifold pressure and engine speed because the
flow at the throttle plate is choked.

Basic formulae to obtain air flow rate for various intake manifold pressure for different engine
speed.
Equating flow through throttle body to engine flow requirement the throttle bore diameter D as a
function of engine rpm can be given as
N
4 Vs vol
D 2
C d Vt
(11)

Where Cd coefficient of discharge through throttle body, vol is volumetric efficiency of engine,
N is engine speed, Vs is displacement volume, and Vt is air velocity through throttle body and is
given by
1
2 P0 P
Vt 1 t
0 1 P0
(12)

Where Po is atmospheric pressure and Pt is intake manifold pressure.


.
Air flow rate m as function of throttle bore diameter D, can be given as
2 1
. D Pt P
1 t

m 2 P0 0
1 P0 P0
4 (13)

6
7

Figure 3:Variation in air flow rate past a throttle, with inlet manifold pressure throttle angle, and
engine speed. 4.7dm3 displacement eight-cylinder engine

Refrences

1. Joh B. Heywood, I ter atio al Co ustio E gi e Fu da e tals M graw-Hill,2011 (page


304-308)

2. Per Carlsson, Linkoping, Flow Through a Throttle Body, February 23, 2007
MATLAB PROGRAMS

Program :1 to find variation of mass flow rate with intake manifold pressure with

various throttle open angle

7
8

Steps to run program:-

1)Type: throttle(D,d,psi0,Cd,gamma)

ref values: D= 0.02m, d=0.004m, psi0=5 ,Cd=0.7,gamma=1.4

2)OUTPUTS: Air flow rate vs Intake manifold pressure

Normalised area vs Throttle angle

function throttle(D,d,psi0,Cd,gamma)
clc
close all

% D=0.02; % enter throttle bore diameter (in m)


% d=0.004; % Enter throttle shaft Diameter (in m)
% psi0=5; % Enter throtle plate angle when tightly closed
against bore (in degree)
% Cd=0.7; % Coefficient Of Discharge
% gamma=1.4; % Enter polytropic/adiabatic constant

CV=(2/(gamma+1))^(gamma/(gamma-1)); %critical value


fprintf('critical value=%0.4f',CV);

a=d/D; % requires later in formulae

P0=1.01325e5; % upstream Pressure (atmospheric pressure in N/m2)


T0=300; % Upstream Temperature (surrounding temperature in
Kelvin)
R=287.05; % in J/kgK

8
9

Pt1=1e4:100:CV*P0; %downstream Pressure(Intake Manifold


Pressure)for Pt/P0 < CV
Pt2 = CV*P0:P0; %downstream Pressure (Intake Manifold
Pressure)for Pt/P0 >CV

psimax=acosd(a*cosd(psi0)); %Wide Open Throttle angle


fprintf('\nWOT=%0.4f',psimax);

% % For various throtle plate angle

for psi=[10,18,21,26,36]

Ath=(pi*(D^2)/4).*((1-
(cosd(psi)./cosd(psi0)))+(2/pi)*((((a./(cosd(psi))).*((cosd(psi)
).^2-a.^2.*(cosd(psi0)).^2).^0.5))-
(cosd(psi).*asin(a.*cosd(psi0)./cosd(psi))/cosd(psi0))+a.*(1-
a.^2).^0.5-asin(a)));
%throttle plate open area

K1=((2/(1.4+1))^((1.4+1)/(2*(1.4-1))))*(1.4^0.5); %for Pt/P0<CV

C=(Pt2./P0);
K2=sqrt((2*1.4/0.4)*(C.^(2/1.4)-C.^(2.4/1.4))); %for Pt/P0>CV

Mth1=(Cd*Ath*P0./(sqrt(R*T0))).*K1;
P_t1=Pt1.*76./1.01325e5; %converting N/m^2
into cm of Hg
M_th1=Mth1.*1e3; %converting Kg/s
into g/s

9
10

Mth2=(Cd*Ath*P0./(sqrt(R.*T0))).*K2;
P_t2=Pt2.*76./1.01325e5; %converting N/m^2
into cm of Hg
M_th2=Mth2.*1e3; %converting Kg/s
into g/s

figure(1)
plot(P_t1,M_th1,'-b',P_t2,M_th2,'-b')
hold on

end

title('Variation in Air Flow Rate Past a Throttle')


xlabel('Intake Manifold Pressure (cm Hg)')
ylabel('Air Flow Rate (g/s)')
grid on

text(4,0.2,'10 deg')
text(4,2,'18 deg')
text(4,3,'21 deg')
text(4,4.8,'26 deg')
text(4,9.4,'36 deg')

%% Variation of throttle area with respect to throttle angle


PSI=psi0:1:psimax;
Ag=(pi*(D^2)/4).*((1-(cosd(PSI)./cosd(psi0)))); %throttle
area neglecting throttle shaft diameter
As=(pi*(D^2)/4).*((1-
(cosd(PSI)./cosd(psi0)))+(2/pi)*((((a./(cosd(PSI))).*((cosd(PSI)
).^2-a.^2.*(cosd(psi0)).^2).^0.5))-

10
11

(cosd(PSI).*asin(a.*cosd(psi0)./cosd(PSI))/cosd(psi0))+a.*(1-
a.^2).^0.5-asin(a)));
%throttle plate open area
figure(2)
plot(PSI,Ag,'-r',PSI,As,'-g')
title('Throttle area estimation')
xlabel('Throttle angle (in degrees)')
ylabel('Normalised area (in meters^2)')
legend('neglecting throttle shaft','including throttle shaft')
grid on

Program 2: to find variation of mass flow rate with intake manifold pressure with

various engine speed

Steps to run program:-

1)Type: throttle1(Cd,Vs,eff)

ref values:- Cd= 0.7, Vs=0.0001m3,eff=0.8

2)OUTPUTS: Air flow rate vs Intake manifold pressure

function throttle1(Cd,Vs,eff)
clc

P0=1.01325e5; %upstream Pressure (atmospheric pressure in N/m2)


T0=300; % Upstream Temperature (surrounding temperature in
Kelvin)
R=287.05; % characterstic gas constant in J/KgK
rho=1.3; % Air density

11
12

% Cd=0.7; % Coefficient Of Discharge


% Vs=0.1e-3; %displacement volume of engine
% eff=0.8; % volumetric efficiency of engine

%% for various RPM

for N=[800,1200,1600,2000,2600,3000,3500,4000]

Pt=1e3:100:0.9*P0; %downstream Pressure(Intake Manifold


Pressure)
P_t=Pt.*76./1.01325e5; %converting N/m^2 into cm of Hg

Vt=sqrt(((2.*P0*1.4)./(1.3*0.4))*((1-Pt./P0).^(0.4/1.4))); %Air
velocity through throttle body
D=sqrt((4.*Vs.*N.*eff)./(2.*1.*Cd.*pi.*Vt));
%throttle open diameter

m=(pi.*D/4).*sqrt((2.*rho.*P0.*(1.4/(1.4-
1))).*((Pt./P0).^(2/1.4)).*((1-(Pt./P0)).^(0.4/1.4))); %mas
flow rate

plot(P_t,m,'-b')
hold on

end

title('Variation in Air Flow Rate Past a Throttle')


xlabel('Intake Manifold Pressure (cm Hg)')
ylabel('Air Flow Rate (g/s)')

12
13

grid on

text(65,5.5,'800rpm')
text(65,6.7,'1200rpm')
text(65,8.4,'2000rpm')
text(65,10.3,'3000rpm')
text(65,11.8,'4000rpm')

13
14

-4
x 10 Throttle area estimation
3
neglecting throttle shaft
including throttle shaft
2.5
Normalised area (in meters 2)

1.5

0.5

-0.5
0 10 20 30 40 50 60 70 80
Throttle angle (in degrees)

Program 3: to find variation of mass flow rate for various throttle open angle

Steps to run program:-

1)Type: throttle2(D,d,psi0,Cd,gamma)

ref values: D= 0.02m, d=0.004m, psi0=5 ,Cd=0.7,gamma=1.4

2)OUTPUTS: Air flow rate vs Throttle open angle

function throttle2(D,d,psi0,Cd,gamma)
clc
close all

% D=0.02; % enter throttle bore diameter (in m)


% d=0.004; % Enter throttle shaft Diameter (in m)

14
15

% psi0=5; % Enter throtle plate angle when tightly closed


against bore (in degree)
% Cd=0.7; % Coefficient Of Discharge
% gamma=1.4; % Enter polytropic/adiabatic constant

a=d/D; % requires later in formulae

a=d/D; % requires later in formulae

P0=1.01325e5; % upstream Pressure (atmospheric pressure in N/m2)


T0=300; % Upstream Temperature (surrounding temperature in
Kelvin)
R=287.05; % in J/kgK

psimax=acosd(a*cosd(psi0)); %Wide open Throttle angle

psi=psi0:psimax;

Ath=(pi*(D^2)/4).*((1-
(cosd(psi)./cosd(psi0)))+(2/pi)*((((a./(cosd(psi))).*((cosd(psi)
).^2-a.^2.*(cosd(psi0)).^2).^0.5))-
(cosd(psi).*asin(a.*cosd(psi0)./cosd(psi))/cosd(psi0))+a.*(1-
a.^2).^0.5-asin(a)));
%throttle plate open area

K1=((2/(1.4+1))^((1.4+1)/(2*(1.4-1))))*(1.4^0.5);

Mth1=(Cd*Ath*P0./(sqrt(R*T0))).*K1; %formula for mass


flow rate
M_th1=Mth1.*1e3; %converting Kg/s
into g/s

15
16

plot(psi,M_th1);
title('Variation in Air Flow Rate Past a Throttle')
xlabel('throtttle open angle (degree)')
ylabel('Air Flow Rate (g/s)')
grid on

Variation in Air Flow Rate Past a Throttle


40

35

30

25
Air Flow Rate (g/s)

20

15

10

-5
0 10 20 30 40 50 60 70 80
throtttle open angle (degree)

Program 3: to find variation of mass flow rate for various coefficient of discharge

Steps to run program:-

1)Type: throttle3(D,d,psi0,gamma)

ref values: D= 0.02m, d=0.004m, psi0=5 ,gamma=1.4

2)OUTPUTS: Air flow rate vs Throttle open angle

16
17

function throttle3(D,d,psi0,gamma)
clc
close all

% D=0.02; % enter throttle bore diameter (in m)


% d=0.004; % Enter throttle shaft Diameter (in m)
% psi0=5; % Enter throtle plate angle when tightly closed
against bore (in degree)
% gamma=1.4; % Enter polytropic/adiabatic constant

a=d/D; % requires later in formulae

P0=1.01325e5; % upstream Pressure (atmospheric pressure in N/m2)


T0=300; % Upstream Temperature (surrounding temperature in
Kelvin)
R=287.05; % in J/kgK

psimax=acosd(a*cosd(psi0)); %Wide Open Throttle angle

psi=psimax;
Cd=0.6:0.00001:1; %coefficient of discharge

Ath=(pi*(D^2)/4).*((1-
(cosd(psi)./cosd(psi0)))+(2/pi)*((((a./(cosd(psi))).*((cosd(psi)
).^2-a.^2.*(cosd(psi0)).^2).^0.5))-
(cosd(psi).*asin(a.*cosd(psi0)./cosd(psi))/cosd(psi0))+a.*(1-
a.^2).^0.5-asin(a)));
%throttle plate open area

17
18

K1=((2/(1.4+1))^((1.4+1)/(2*(1.4-1))))*(1.4^0.5); %for Pt/P0<CV


Mth1=(Cd*Ath*P0./(sqrt(R*T0))).*K1;
M_th1=Mth1.*1e3; %converting Kg/s
into g/s

plot(Cd,M_th1,'-b')
title('Variation in Air Flow Rate Past a Throttle')
xlabel('Coefficient of Discharge (Cd)')
ylabel('Air Flow Rate (g/s)')
grid on

Variation in Air Flow Rate Past a Throttle


46

44

42

40
Air Flow Rate (g/s)

38

36

34

32

30

28

26
0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1
Coefficient of Discharge (Cd)

18
1

Poppet Valve Geometry and Timing

Figure shows the main geometric parameters of a poppet valve head and seat. Figure shows the
proportions of typical inlet and exhaust-valves and ports relative to the valve inner seat diameter
D. Typical valve head sizes for different shaped combustion chambers in terms of cylinder
bore B are given in Table . Each of these chamber shapes allow higher maximum air flows for a
given cylinder displacement.

Figure 1

Typical valve timing, valve-lift profiles, and valve open areas for a four-stroke cycle spark-
ignition engine are shown in Figure There is no universally accepted criterion for defining valve
timing points. Some are based upon a specific lift criterion.
2

Figure 2

Table-1
3

The instantaneous valve flow area depends on valve lift and the geometric details of the valve
head, seat, and steam. There are three separate stages to the details of the flow area development
as valve lift increases, as shown in Fig. 6-14b.

For low valve lifts, the minimum flow area corresponds to a frustrum of a right circular cone
where the conical face between the valve and the seat, which is perpendicular to the seat, defines
the flow area. For this stage :

Formula-1:

> > 0,


= + ,

where B is the valve seat angle, Lv is the valve lift, Dv is the valve head diameter and w is the
seat width.

For the second stage, the minimum area is still the slant surface of a frustrum of a right circular
cone, but this surface is no longer perpendicular to the valve seat. The base angle of the cone
increases from (90 B) toward that of a cylinder, 90. For this stage :

Formula-2 :


+ > ,

= [ + ]
4

where Dp is the port diameter, Ds is the valve stem diameter and Dm is the mean seat diameter
(Dv -w).

Figure-3

Finally, when the valve lift is sufficiently large, the minimum flow area is no longer between the
valve head and seat ; it is the port flow area minus the sectional area of the valve stem.
5

Thus for

Formula-3


> + , = ( )

Intake and exhaust valve open areas corresponding to a typical valve-lift profile are plotted
versus camshaft angle in Fig. 6-14c. These three different flow regimes are indicated. The
maximum valve lift is normally about 12 percent of the cylinder bore.

The effect of valve geometry and timing on air flow can be illustrated conceptually by dividing
the rate of change of cylinder volume by the instantaneous minimum valve flow area to obtain
a pseudo flow velocity for each valve :

Formula-4:


= =

where V is the cylinder volume, B is the cylinder bore,


s is the distance between the wrist pin and crank axis
and Am is the valve area given by Eqs (6.7), (6.8) or (6.9).

Instantaneous pseudo flow velocity profiles for the exhaust and intake strokes of a four-stroke
four-cylinder engine are shown in Fig. 6-15.
6

Figure-4 :

Note the appearance of two peaks in the pseudo flow velocity for both the exhaust and the intake
strokes. The broad peaks occuring at maximum piston velocity reflect the fact that valve flow
area is constant at this point. The peaks close to TDC result from the exhaust valve closing and
intake valve opening profiles. The peak at the end of the exhaust stroke is important since it
indicates a gigh pressure drop across the valve at this point, which will result in higher trapped
residual mass. The magnitude of this exhaust stroke pseudo velocity peak depends strongly on
the timing of exhaust valve closing.

The pseudo velocity peak at the start of the intake stroke in much less important.
7

That the pseudo velocities early in the exhaust stroke and late in the intake stroke are low
indicates that phenomena other than quasi-steady flow governe the flow rate. These are the
periods when exhaust blowdown and ram and tuning effects in the intake are most important.

Plotting in Matlab:

Cam profile selection was based on SHM equation given by following method.

Simple harmonic motion curve is widely used since it is simple to design. The curve is the
projection of a circle about the cam rotation axis as shown in the figure. The equations relating
the follower displacement velocity and acceleration to the cam rotation angle are:

In figure below the displacement, velocity and acceleration curves are shown. The maximum
velocity and acceleration values given by equations:

vmax= , amax=
8

Plotting instantaneous valve lift curve:

Program:

%instantaneous valve lift curve


x=0:0.001:228;%duration of valve lift
l=5*(1-cos((2*3.1415*x)/228));%simple harmonic equation for cam profile
plot(x,l,'LineWidth',1.5);
9

Output:

Plotting instantaneous inlet area opening:

Code:

%Bore diameter was taken as 84mm,corresponding valve head dia=36.12mm and


%other calculations like Ds,Dp,w, was selected 30 degrees and D were
obtained from empirical relations from Heywood.Empirical relations also given
in figure 2 and table 1 of document.
for(x=0:0.01:50.95)%stage 1

p=5*(1-cos((3.1415.*x)/114));%lift calculation instantaneous


10

a=3.14.*p*cos(3.14/6)*(36.3-6.567+5*sin(3.14/3));%area calculation
instantaneous
plot(x,a,'LineWidth',1.5)
grid on;
hold on;
end%end of first stage

for(x=50.95:0.01:100)%second stage

p=5*(1-cos((3.1415.*x)/114));%lift calculation instantaneous


a=3.14*30.715*((p-(3.2835*tan(3.14/6)))^2+3.285^2)^0.5;%area calculation
instantaneous
plot(x,a,'LineWidth',1.5)
grid on;
hold on;
end%end of second stage

for(x=100:0.01:114)%stage three
plot(x,810,'LineWidth',1.5)%full open flow area of 810mm^2 for 14 degrees
grid on;
hold on;
end

for(x=114:0.01:128)%reverse stage three


plot(x,810,'LineWidth',1.5)
grid on;
hold on;
end

for(x=177.05:-0.01:128)%reverse stage two

p=5*(1-cos((3.1415.*x)/114));
a=3.14*30.715*((p-(3.2835*tan(3.14/6)))^2+3.285^2)^0.5;
plot(x,a,'LineWidth',1.5)
grid on;
hold on;
end
11

for(x=228:-0.01:177.05)%reverse stage one

p=5*(1-cos((3.1415.*x)/114));
a=3.14.*p*cos(3.14/6)*(36.3-6.567+5*sin(3.14/3));
plot(x,a,'LineWidth',1.5)
grid on;
hold on;
end
Output:

References:

1. Internal Combustion Engine Fundamentals (Mcgraw Hill Series in Mechanical


Engineering) , 1 May 1988. by John Heywood (Author)page221-238.

2.Cam design and manufacture by Preben W. Jenson,Chapter 2.


1

Valve Flow Model


Figure shows the main geometric parameters of a poppet valve head and seat. Figure shows the
proportions of typical inlet and exhaust-valves and ports relative to the valve inner seat diameter
D. Typical valve head sizes for different shaped combustion chambers in terms of cylinder
bore B are given in Table . Each of these chamber shapes allow higher maximum air flows for a
given cylinder displacement.

Figure 1

Typical valve timing, valve-lift profiles, and valve open areas for a four-stroke cycle spark-
ignition engine are shown in Figure There is no universally accepted criterion for defining valve
timing points. Some are based upon a specific lift criterion.
2

Figure 2

Table-1
3

The instantaneous valve flow area depends on valve lift and the geometric details of the valve
head, seat, and steam. There are three separate stages to the details of the flow area development
as valve lift increases, as shown in Fig. 6-14b.

For low valve lifts, the minimum flow area corresponds to a frustrum of a right circular cone
where the conical face between the valve and the seat, which is perpendicular to the seat, defines
the flow area. For this stage :

Formula-1:

> > 0,


= + ,

where B is the valve seat angle, Lv is the valve lift, Dv is the valve head diameter and w is the
seat width.

For the second stage, the minimum area is still the slant surface of a frustrum of a right circular
cone, but this surface is no longer perpendicular to the valve seat. The base angle of the cone
increases from (90 B) toward that of a cylinder, 90. For this stage :

Formula-2 :


+ > ,

= [ + ]

where Dp is the port diameter, Ds is the valve stem diameter and Dm is the mean seat diameter
(Dv -w).
4

Figure-3 :
5

Finally, when the valve lift is sufficiently large, the minimum flow area is no longer between the
valve head and seat ; it is the port flow area minus the sectional area of the valve stem.

Thus for


> + , = ( )

Formula-3

Intake and exhaust valve open areas corresponding to a typical valve-lift profile are plotted
versus camshaft angle in Fig. 6-14c. These three different flow regimes are indicated. The
maximum valve lift is normally about 12 percent of the cylinder bore.

The effect of valve geometry and timing on air flow can be illustrated conceptually by dividing
the rate of change of cylinder volume by the instantaneous minimum valve flow area to obtain
a pseudo flow velocity for each valve :

Formula-4:


= =

where V is the cylinder volume, B is the cylinder bore,


s is the distance between the wrist pin and crank axis
and Am is the valve area given by Eqs (6.7), (6.8) or (6.9).

Formula-5:Mass flow rate through valve

1
2

P 1 P2 P
m C D AT P1 2 2
P1 RT1 P1 P1

Where C D is coefficient of discharge, AT is valve flow area,


6

P1 and P2 are atmosphere and cylinder pressure

Plotting in Matlab:

Cam profile selection was based on SHM equation given by following method.

Simple harmonic motion curve is widely used since it is simple to design. The curve is the
projection of a circle about the cam rotation axis as shown in the figure. The equations relating
the follower displacement velocity and acceleration to the cam rotation angle are:

In figure below the displacement, velocity and acceleration curves are shown. The maximum
velocity and acceleration values given by equations:

vmax= , amax=
7

Plotting instantaneous valve lift curve:

Program:
%instantaneous valve lift curve
x=0:0.001:228;%duration of valve lift
l=5*(1-cos((2*3.1415*x)/228));%simple harmonic equation for cam profile
plot(x,l,'LineWidth',1.5);

Output:
8

Plotting instantaneous inlet area opening:

Code:
%Bore diameter was taken as 84mm,corresponding valve head dia=36.12mm and
%other calculations like Ds,Dp,w, was selected 30 degrees and D were
obtained from empirical relations from Heywood.Empirical relations also given
in figure 2 and table 1 of document.
for(x=0:0.01:50.95)%stage 1

p=5*(1-cos((3.1415.*x)/114));%lift calculation instantaneous


a=3.14.*p*cos(3.14/6)*(36.3-6.567+5*sin(3.14/3));%area calculation
instantaneous
plot(x,a,'LineWidth',1.5)
grid on;
hold on;
9

end%end of first stage

for(x=50.95:0.01:100)%second stage

p=5*(1-cos((3.1415.*x)/114));%lift calculation instantaneous


a=3.14*30.715*((p-(3.2835*tan(3.14/6)))^2+3.285^2)^0.5;%area calculation
instantaneous
plot(x,a,'LineWidth',1.5)
grid on;
hold on;
end%end of second stage

for(x=100:0.01:114)%stage three
plot(x,810,'LineWidth',1.5)%full open flow area of 810mm^2 for 14 degrees
grid on;
hold on;
end

for(x=114:0.01:128)%reverse stage three


plot(x,810,'LineWidth',1.5)
grid on;
hold on;
end

for(x=177.05:-0.01:128)%reverse stage two

p=5*(1-cos((3.1415.*x)/114));
a=3.14*30.715*((p-(3.2835*tan(3.14/6)))^2+3.285^2)^0.5;
plot(x,a,'LineWidth',1.5)
grid on;
hold on;
end

for(x=228:-0.01:177.05)%reverse stage one

p=5*(1-cos((3.1415.*x)/114));
a=3.14.*p*cos(3.14/6)*(36.3-6.567+5*sin(3.14/3));
plot(x,a,'LineWidth',1.5)
grid on;
hold on;
end
Output:
10

Plotting Mass Flow Rate vs Crank Angle


%Programme for mass flow rate vs crank angle
data= xlsread('p-v data_engine 2.xlsx'); %Import data for pressure and crank
angle
ca= data(:,1); %saves the values of crank angle
p= data(:,2); %saves the values of pressure
m=0.5*8.324*1.155*sqrt((1/12300)*(((p./155148).^1.714286)-
((p./1155148).^1.4286)));
%Cd=0.5 ,P1=1.155,y=1.4,R=287,T1=300k
plot(ca,m,'linewidth',1.5) %plots the graph
grid on;
xlabel('crank angle theta(in degrees)');
ylabel('mass flow rate(in kg/s)');
title('mass flow rate of a 4-stroke reciprocating engine');
legend('pressure','location','northeast');
11

Output:

References:

1. Internal Combustion Engine Fundamentals (Mcgraw Hill Series in Mechanical


Engineering) , 1 May 1988. by John Heywood (Author)page221-238.

2.Cam design and manufacture by Preben W. Jenson,Chapter 2.


Modeling of NOx formation in IC Engines

Vishwajit Doijad

RollNo.12 ; MIS No.121195011


Academic Student ,Automotive Technology,COEP

Copyright 2009 SAE International

ABSTRACT burned gas decrease due to mixing with cool bulk gas
and expansion of the charge. Fuel-Air equivalence ratio
From last two decades, the work done by John is another important factor influencing NOx formation. As
B.Heywood and Dr.Gnter P. Merker has strongly the equivalence ratio becomes leaner, NO and NOx
contributed to improve emission quality to attend decrease significantly as expected. NO2 however shows
engines and environment regulations evolutions. an opposite trend to that of NO that causes the
NO2/NOx ratio to increase at leaner conditions. The
In this context, I considered studies made by John NO2peaks at an equivalence ratio near 0.25. Leaner
B.Heywood and Dr.Gnter P. Merker about pollutant. equivalence ratio is indicative flower loads and lower
This context presents the statistical modeling studies bulk gas temperatures that are conducive to the
which allow predicting these pollutants. formation of NO2. Advancing injection timing or
increasing injection pressure improves combustion
INTRODUCTION efficiency raises combustion temperature. In general,
higher combustion temperatures lead to higher NOx
Considering the heterogeneous nature of fuel-air mixture formation. Addition of diluents to the engine intake air is
in diesel engines NOx and Particulate matter (PM) are considered as an effective mean to reduce the NO
important emissions. Continuous efforts are being made formation rate and hence the exhaust NOx levels. The
to minimize the quantities of these two pollutants from effect is primarily one of reducing the peak flame
the diesel engine exhaust. Modelling and gas sampling temperature, which is the driving factor for NOx
studies have been done with indirect injection(IDI) diesel formation. Diluents such as N2, CO2 and exhaust gas
engines, which suggest that pre chamber is the were added to the intake air of direct injection (DI)
prominent location for formation of nitrogen oxides. engine to study their effect on NOx reduction. NOx
There are a number of potential mechanisms emissions comprise of NO and NO2. The NO2 is formed
responsible for NO in combustion processes. The via NO molecule. Therefore, the modelling of NOx
relative importance of these different mechanisms is formation is most often reduced to studying the
strongly affected by the temperature, fuel-air formation of NO. It is widely accepted that in diesel
equivalence ratio, pressure, flame conditions, residence engines the major portion of NO is formed via thermal
time and concentrations of key reacting species. Rapid path.
NOx formation begins after the start of heat release.
Shortly after the end of heat release, the period of rapid KINETICS OF NOX FORMATION:
NOx formation ends because temperatures of the

DO NOT TYPE IN THIS FOOTER SECTION ON PAGE 1. TEXT WILL BE INSERTED INTO THIS AREA BY SAE STAFF UPON RECEIPT OF THE FINAL
APPROVED MANUSCRIPT AT SAE INTERNATIONAL.
Nitric oxide (NO) and nitrogen dioxide (NO,) are usually
grouped together as NOx, emissions. Nitric oxide is the
predominant oxide of nitrogen produced inside the
engine cylinder. The principal source of NO is the
oxidation of atmospheric (molecular) nitrogen. However,
if the fuel contains significant nitrogen the oxidation of
the fuel nitrogen-containing compounds is an additional
source of NO. Gasoline contains negligible amounts of
nitrogen; although diesel fuel contain more nitrogen,
current levels are not significant. NO formation was first
described by Zeldovich (1946). This extended Zeldovich
mechanism consists of three elementary reactions with
experimentally determined speed constants ki.

Tab.1: Speed coefficients for the forward reaction of the


[Ref1] Zeldovich mechanism) {Ref1}

Let + Index: forward reaction We know by the speed coefficients for the forward
reactions that NO formation via reaction (1) progresses
- Index: reverse reaction much more slowly than via reactions (2) and (3).
Because the reaction speed of the forward reactions (2)
For the NO formation rate, we obtain with reaction and (3) is higher than that of reaction(1) by several
equations (1) to (3) _ decimal powers, the atomic nitrogen formed in the first
reaction step is immediately further converted to NO in
the second and third steps. The concentration of atomic
nitrogen thus remains constant after a short initial phase.
The concentration of [N] can thus be assumed to be
quasi-steady therefore_

[Ref2]
[Rrf1]
And for the temporal change of nitrogen atom The NO formation rate then becomes
concentrations follows

Where,
[Ref2]
+
Speed coefficients for the forward reaction (k i or ki, r) of
the Zeldovich mechanism in following table1, NO forms in both the flame front and the post flame
gases. In engines however, combustion occurs at high
pressure so the flame reaction zone is extremely thin (-
0.1 mm) and residence time within this zone is short.
Also the cylinder pressure rises during most of the
combustion process, so burned gases produced early in
the combustion process are compressed to a higher
temperature than they reached immediately after
combustion. Thus, NO formation in the postflame gases
almost always dominates any flame-front-produced NO.
it is, therefore, appropriate to assume that the
combustion and NO formation processes are decoupled
and to approximate the concentrations of O2, O,OH,H
and N, by their equilibrium values at the local pressure
and equilibrium temperature TO introduce this an engine), this initial NO formation rate peaks at the
equilibrium assumption it is convenient to use the stoichiometric composition, and decreases rapidly as the
notation mixture becomes leaner or richer.
+ -
R1=K1 [O]e[N]e = K1 [NO]e[N]e
+ -
R2= K2 [N]e[O2]e = K2 [NO]e[O]e
+ -
R3= K3 [N]e[OH]e = K2 [NO]e[H]e
The above eqn. reduces to,

Where [ ] denotes equilibrium concentration for the one-


way equilibrium rate.The values of R1, R2 and R1/
(R2+R3) are given in the table below.

The strong temperature dependence of the NO


formation rate can be demonstrated by considering the
initial value of d[NO]/dt when [NO]/[NO]e << 1. Then
above eq. becomes
FORMATION TIME: -

The equilibrium oxygen atom concentration is given by- A characteristic time for the NO formation process can
be defined by,

[Ref2]
Where Kp (O) is the equilibrium constant for the reaction
and
[Ref2]

[Ref2]
Then the initial NO formation rate can be written as

This is the mathematical governing model of NOx


Formation. The strong dependence of d[NO]/dt on
temperature in the exponential term is evident. High
temperatures and high oxygen concentrations result in
high NO formation rates. The Figure shows the NO
formation rate as a function of gas temperature and
fuel/air equivalence ratio in post flame gases. Also
shown is the adiabatic flame temperature attained by a
fuel-air mixture initially at 700 K at a constant pressure of
15 atm. For adiabatic constant-pressure combustion (an
appropriate model for each element of fuel that burns in
NO FORMATION IN SPARK-IGNITION ENGINES

In conventional spark-ignition engines the fuel and air


(and any recycled exhaust) are mixed together in the
engine intake system, and vigorous mixing with the
residual gas within the cylinder occurs during the intake
process. Thus the fuel/air ratio and the amount of
diluents (residual gas plus any recycled exhaust) is
approximately uniform throughout the charge within the
cylinder during combustion.

Fig: - (a) Measured cylinder pressure p and calculated


mass fraction burned Xb

Fig (a): NO and NO2 concentrations in SI engine as


function of air/fuel ratio

Fig: - (b) Calculated temperature of unburned gas and


burned gas in early- and late burning elements

Fig.(b) NO2 as percent of total NOx in diesel exhaust as


function of load and speed. Fig: - (c) Calculated NO concentrations in early- and
late-burning elements For rate controlled model and at
equilibrium
The fig. shows examples of NO and NO emissions data The NO concentration gradient across gas in the engine
from a spark ignition and a diesel engine. The maximum cylinder, due to the temperature gradient, has been
value for the ratio (NO2/NO) for the SI engine is 2 demonstrated sampling techniques and using
percent at an equivalence ratio of about 0.85. For the measurements of the chemiluminescent radiation from
diesel this ratio is higher and is highest at light load and the reaction,
depends on engine speed. NO +O2-- NO2.
For determining the local NO concentration with the help
of figure NO concentration data as a function of crank
angle taken by Lavoie two different windows in the
cylinder head of a specially constructed L head engine
where each window was a different distance from the
spark plug.

Fig: - Variation of exhaust NO concentration


With A/F and fuel/air equivalence ratio.

SPARK TIMING

Spark timing significantly affects NO emission level.


advancing the timing so that combustion occurs earlier in
the cycle increases the peak cylinder pressure (because
more fuel is burned before TC and the peak pressure
Fig: - Spectroscopic ally measured NO concentrations moves closer to TC where the cylinder volume is
through two windows W3 and W2 in special L-head SI smaller) retarding the timing decreases the peak cylinder
engine (W2 is closer to spark than W3).The asterisks pressure (because more of the fuel burns after TC).
mark estimated initial conditions and flame arrival times. Higher peak cylinder pressures result in higher peak
The dashed lines are calculated rate limited burned gas temperatures, and hence higher NO
concentrations for parts of charge burning at these flame formation rates. For lower peak cylinder pressures lower
arrival times with zero initial NO concentration. NO formation rates result.

The figure shows typical NO emission data for a spark-


EQUIVALENCE RATIO ignition engine as a function of spark timing. NO
emission levels steadily decrease as spark timing is
Fig shows the effect of variations in the fuel air retarded from MBT timing and moved closer to TC.
equivalence ratio on NO emissions. Maximum burned Since exact determination of MBT timing is difficult (and
gas temperatures occur at equivalence ratio = 1.1 not critical for fuel consumption and power where the
however at this equivalence ratio oxygen concentrations variation with timing around MBT is modest) there is
are low. As the mixture is enriched burned gas always considerable uncertainty in NO emissions at
temperatures fall and as the mixture is leaned out MBT timing.
increasing oxygen concentration initially offsets the
falling gas temperatures and NO emissions peak at 0.9.
Detailed predictions of NO concentrations in the burned
gases suggest that the concentration versus time
histories under fuel-lean conditions is different in
character from those for fuel-rich conditions. In lean
mixtures NO concentrations freeze early in the
expansion process and little NO decomposition occurs.
In rich mixtures, substantial NO decomposition occurs
from the peak concentrations present when the cylinder
pressure is a maximum Thus in lean mixtures, gas
conditions at the time of peak pressure are especially
significant.
Fig: - Concentrations of soot, NO, and other combustion
Fig: -Variation of exhaust NO concentration with spark
product species measured at outer edge of bowl in
retard.
piston combustion chamber of quiescent DI diesel with
rapid sampling valve.
NOX FORMATION IN COMPRESSION-IGNITION
ENGINES The critical time period is when gas temperatures are at
a maximum burned. Mixture which burns early in the
The kinetic mechanisms for NO and NO, formation combustion process is especially important since it is
described in above Sections and the assumptions made compressed to a higher temperature increasing the NO
regarding equilibration of species in the C-0-H system formation rate as combustion proceeds and cylinder
apply to diesels as well as to spark-ignition engines. The pressure increases. After the time of peak pressure
critical difference of course is that injection of fuel into burned gas temperatures decrease as the cylinder
the cylinder occurs just before combustion starts and gases expand. The decreasing temperature due to
that non uniform burned gas temperature and expansion and due to mixing of High-temperature gas
composition result from this non uniform fuel distribution with air or cooler burned gas freezes the NO chemistry.
during combustion. The fuel-air mixing and combustion This second effect (which occurs only in the diesel)
processes are extremely complex. During the mixing, means that freezing occurs more rapidly in the diesel
Controlled combustion phase the burning mixture is than in the spark-ignition engine, and much less
likely to be closer to stoichiometric (the flame structure is decomposition of the NO occurs. The above description
that of a turbulent, though unsteady diffusion flame). is supported by the NO concentration data obtained from
However, throughout the combustion process mixing experiments where gas was sampled from within the
between already burned gases, air and lean and rich cylinder of normally operating diesel engines with special
unburned fuel vapour-air mixture occurs changing the gas-sampling valves and analysed. Figure shows time
composition of any gas elements that burned at a histories of major species concentrations, through the
particular equivalence ratio. In addition to these combustion process determined with a rapid-acting
composition (and hence temperature) changes due to sampling valve (1 ms open time) in a quiescent direct-
mixing, temperature changes due to compression and injection diesel engine.
expansion occur as the cylinder pressure rises and falls.
The critical equivalence ratio for NO formation in high- EQUIVALENCE RATIO:-
temperature high-pressure burned gases typical of
engines is close to stoichiometric. Figure shows the The effect of the overall equivalence ratio on NO
initial NO formation rate in combustion products formed concentrations is shown in the below Fig. At high load
by burning a mixture of a typical hydrocarbon fuel with with higher peak pressures and hence temperatures and
air (initially at 700 K at a constant pressure of 15 atm). larger regions of close-to-stoichiometric burned gas NO
NO formation rates are within a factor of 2 of the levels increase. NO2 is 10 to 20 percent of total NOx.
maximum value for 0.85< Equ. Ratio< 1.1. Though NO levels decrease with a decreasing overall
equivalence ratio they do so much less rapidly than do
spark-ignition engine NO emissions due to the non-
uniform fuel distribution in the diesel. Though the amount
of fuel injected decreases proportionally as the overall
equivalence ratio is decreased much of the fuel still
burns close to stoichiometric. Thus NO emissions should
be roughly proportional to the mass of fuel injected.
Fig: -Exhaust NOx and NO concentrations as a function
of overall equivalence ratio

REFERENCES

[Ref1]: Simulating Combustion by Gunter P Merker


Christian Schwarz Gunnar Stiesch Frank Otto;
Springer Publication;Ed2006;Page No.132 -140.
[Ref2]:Internal Combustion Engine Fundamental by
John B Heywood; McGraw-Hill Publication;
Ed1988;Page No.572-586.
[Ref3]:Modeling of Diesel Combustion by P.A.
Lakshminarayanan. Yogesh V. Aghav;
Springer Publication;Ed2010;Page No.189-201
NITROGEN OXIDES

Nitrogen oxides are important air pollutants, the primary anthropogenic source of
which is combustion. Motor vehicles account for a large fraction of the nitrogen oxide
emissions, but stationary combustion sources ranging from electric power generating stations
to gas-fired cooking stoves also release nitrogen oxides.
Both nitric oxide (NO), and nitrogen dioxide (NO2), are produced in combustion, but
the vast majority of nitrogen oxides are emitted as NO. Because NO is converted to NO2 in
the atmosphere, emissions of both species frequently are lumped together with the
designation NOx. When NOx emissions are presented in mass units, the mass of NOx is
calculated as if all the NO had been converted to NO2. Because NO is the predominant NOx,
species found in combustion, we shall concentrate on it in this section.
Nitric oxide is formed both from atmospheric nitrogen, N2, and from nitrogen
contained in some fuels. The latter source depends on the fuel composition and is not
important for fuels with low nitrogen contents but is a major source of NO, in coal
combustion. Nitric oxide can be formed, however, when any fuel is burned in air because of
the high-temperature oxidation of N2. We begin our discussion with the fixation of
atmospheric nitrogen.

Thermal Fixation of Atmospheric Nitrogen

The formation of NO by oxidation of atmospheric nitrogen can be expressed in terms of the


overall reaction

N + O NO

This reaction is highly endothermic. As a result, the equilibrium concentration of NO


is high at the very high temperatures encountered near stoichiometric combustion and
decreases rapidly away from that point.
Even though we express the overall reaction as above, the direct reaction of N2 with
O2 is too slow to account for significant NO formation. Free oxygen atoms, produced in
flames by dissociation of O2 or by radical attack on O2, attack nitrogen molecules and begin a
simple chain mechanism that was first postulated by Zeldovich.
N + O NO + N
N + O NO + O

The concentration of O2 is low in fuel-rich combustion, so reaction 2 is less important than in


fuel-lean combustion. Reaction with the hydroxyl radical eventually becomes the major sink
for N
N + OH NO + O
The rate constants for the so-called extended Zeldovich mechanism are given as

k+ = . /T

k = . /T

k+ = . /T

k = . /T

k+ = . /T

k = . /T
-(1)

The high activation energy of reaction 1, resulting from its essential function of
breaking the strong N2 triple bond, makes this the rate-limiting step of the Zeldovich
mechanism.
Due to the high activation energy, NO production by this mechanism proceeds at a
slower rate than the oxidation of the fuel constituents and is extremely temperature sensitive.
The production of atomic oxygen required for the first reaction is also highly temperature
sensitive.
To understand the rate of NO formation, let us examine the rate equations
corresponding to the mechanism of reactions 1-3. For example, the net rates of formation of
NO and N are

R = k + [N ] [O] k [N] [NO] + k + [N] [O ] k [NO] [O] + k + [N] [OH]


k [NO] [H]

R = k + [N ] [O] k [N] [NO] + k + [N] [O ] k [NO] [O] + k + [N] [OH]


k [NO] [H]
The concentrations of O, H, and OH are required for calculation of the N and NO formation
rates. The high activation energy of the initial N2 attack allows us to make an important
simplification. Since the reaction rate is fast only at the highest temperatures, most of the
reaction takes place after the combustion reactions are complete and before significant heat is
transferred from the flame. It is a reasonable first approximation, therefore, to assume that the
O, H, and OH radicals are present in their equilibrium concentrations. This suggests a
simplification proposed by Lavoie.
At thermodynamic equilibrium, we may write

k + [N ]e [O]e = k [N]e [NO]e

We may define the equilibrium, one-way rate of reaction as

R = k + [N ]e [O]e = k [N]e [NO]e

R = k + [N]e [O ]e = k [NO]e [O]e

R = k + [N]e [OH]e = k [NO]e [H]e - ( 2)

Further defining the quantities,


[ ]
=
[NO]e
[ ]
=
[N]e
-(3)
the rate equations may now be expressed in the abbreviated form as

R =R R + R R + R R
R = R R R + R R +R

We must determine the N atom concentration if we are to calculate the rate of NO formation.
Since the activation energy for oxidation of the nitrogen atom is small and for fuel-lean
conditions, the reaction involves O2, which is a major component of the gas. The free
nitrogen atoms are consumed as rapidly as they are generated, establishing a quasi steady
state. Setting the left-hand side of R equal to zero and solving for the steady state nitrogen
atom concentration, we find
R +R +R +
= =
R +R +R +
Where,

R
=
R +R
-(4)
Substituting value of into R yields a rate equation for NO formation in terms of .
And known quantities,
R
R =
+k
-(5)
The initial NO formation rate (at = 0) is twice the rate of reaction 1. The sharp peak near
stoichiometric is due to the high flame temperatures.

Assuming that there is no NO present initially (i.e., = 0 at t = 0), the result is


ln + + ln =

-(6)
Where the characteristic time for NO formation is :

[NO]e
=
R
-(7)
It is apparent that corresponds to the time that would be required for NO to reach
the equilibrium level if the reaction continued at its initial rate and were not slowed by the
reverse reactions.
Two major assumptions have been made in the derivation of equation (5).
A quasi steady state for the nitrogen atom concentration
Equilibrium concentrations for the O, H, and OH radicals.
The validity of the first assumption can readily be examined.
Flow chart

Select fuel and input values for inlet air


pressure & temperature,A/F Ratio
Compression ratio,Calorific value,
Residence time

Stoichimetric and fuel lean Rich combustion


combustion

Calculate adiabatic flame


temperature

Finding NO concentration using


the data obtained from
Calculate temperature dependent rate
Heywood(Valid only for petrol
constants and find equilibrium
engine at T0=290K, P0=1atm ,
concentrations
r=10)

Calculate characteristic time,NO


concentration and NO formation
rate
Procedure for finding NO concentration
Inputs required:
a, b = fuel of type CaHb
Lambda = A/F ratio
r = Compression ratio
T0, P0 = Inlet temperature (K) and pressure (atm).
HCV = higher calorific value in MJ/kg
tr = residence time

Now, value of gamma () for air is 1.4


So, temperature of air after compression is
0
=

Where is pressure of air after compression :


=
Now, to calculate the flame temperature (T) after combustion, method of enthalpy balance is
used and enthalpies of gases in the exhaust is calculated by using equations of combustion at
various A/F ratio.


[ ]= +

Values of and are given as:


Table 1- Values of &

Gas

CO 44.319 0.00730
O 30.504 0.00349
N 29.231 0.00307
H O 32.476 0.00862

From the flame temperature, equilibrium constants and concentration of gases is calculated
using equations 1, 2, 5.
Characteristic time is then calculated using equation 7.
Then finally, concentration of NO is calculated by finding from 6.
Program execution

Two separate programs are made for getting Nitric Oxide(NO) concentration and for
getting various plots
The file names are NO.m and plot123.m
User needs to enter the values of a, b, lambda, r, T0, P0, HCV, tr.
Where ,
a, b = fuel of type CaHb
Lambda = A/F ratio
R = Compression ratio
T0, P0 = Inlet temperature (K) and pressure (atm).
HCV = higher calorific value in MJ/kg
tr = residence time
Standard values for ready reference:
NO (8, 18, 1, 10, 290, 1, 46.536, 0.005)
Plots :

Figure 1- Variation of NO concentration with Air fuel ratio

Figure 1 shows influence of variation in air-fuel ratio on NO emissions. In this plot, it


is clear that NO emissions are peaking at air/fuel ratio around 1.1


Figure 2- Variation of with

Figure 2 describes the approach to equilibrium. It is apparent that corresponds to


the time that would be required for NO to reach the equilibrium level if the reaction
continued at its initial rate and were not slowed by the reverse reactions.
Matlab Program
% Program for studying variation of NO concentration with temperature and
% A/F ratio

clc
clear all
lambda= 0.85:0.005:200;

for i= 1:1: lambda(end)

%Constants for calculating specific heats at constant pressure


a1 = 44.319; % for CO2
b1 = 0.00730;

a2 = 32.477; % for O2
b2 = 0.00862;

a3 = 29.231; % for N2
b3 = 0.00307;

a4 = 30.504; % for H2O


b4 = 0.00349;

a = 8; % (For petrol a = 8 ,b = 18)


b = 18; % (For petrol a = 12 ,b = 18)
r = 10; % Considering Compression ratio = 10
P0 = 1; % Considering ambient pressure = 1 atm
T0 = 290; % Considering ambient temperature = 290K
HCV = 46.536; % Higher Calorific value of petrol in MJ/kg
tr = 0.005; % Residence time-generally taken as 0.005 seconds

% Temperature and pressure after compression


Gamma = 1.4; % Gamma = 1.4 for air
Ti = T0 * r^((Gamma-1)/Gamma);
p1 = r * P0;

% Matching andplotting the maximum concentrations at lambda=1.07


if lambda(i) >= 1.07
if lambda(i) == 1.162
lambda(i) = 1.1620000001;
end

% Concentrations of gases O2, N2, H2O in the engine exhaust

yO2 = (((a + (b/4)) * (lambda(i) - 1)))/(a + b + 3.78 * (a + (b/4)) *


lambda(i) + (a + (b/4)) * (lambda(i) - 1));
yN2 = (3.78 * (a + (b/4)) * lambda(i))/(a + b + 3.78 * (a + (b/4)) *
lambda(i) + (a + (b/4)) * (lambda(i) - 1));
yH2O = (b/2)/(a + b + 3.78 * (a + (b/4)) * lambda(i) + (a + (b/4)) *
(lambda(i) - 1));

% Calculation of Adiabatic flame temperature


% i.e Temperature after combustion

T = -(200*a*a1 - 200*a*a4 + 100*a2*b - 50*a4*b - 200*(T0^2*a^2*b1^2 +


(189*T0^2*a^2*b1*b3*lambda(i))/50 + T0^2*a^2*b1*b4*lambda(i) -
2*T0^2*a^2*b1*b4 - (189*T0^2*a^2*b3*b4*lambda(i))/50 -
T0^2*a^2*b4^2*lambda(i) + T0^2*a^2*b4^2 + T0^2*a*b*b1*b2 +
(189*T0^2*a*b*b1*b3*lambda(i))/200 + (T0^2*a*b*b1*b4*lambda(i))/4 -
(T0^2*a*b*b1*b4)/2 + (189*T0^2*a*b*b2*b3*lambda(i))/100 +
(T0^2*a*b*b2*b4*lambda(i))/2 - T0^2*a*b*b2*b4 -
(189*T0^2*a*b*b3*b4*lambda(i))/100 - (T0^2*a*b*b4^2*lambda(i))/2 +
(T0^2*a*b*b4^2)/2 + (T0^2*b^2*b2^2)/4 + (189*T0^2*b^2*b2*b3*lambda(i))/400
+ (T0^2*b^2*b2*b4*lambda(i))/8 - (T0^2*b^2*b2*b4)/4 -
(189*T0^2*b^2*b3*b4*lambda(i))/800 - (T0^2*b^2*b4^2*lambda(i))/16 +
(T0^2*b^2*b4^2)/16 + 2*T0*a^2*a1*b1 + (189*T0*a^2*a1*b3*lambda(i))/25 +
2*T0*a^2*a1*b4*lambda(i) - 2*T0*a^2*a1*b4 - 2*T0*a^2*a4*b1 -
(189*T0*a^2*a4*b3*lambda(i))/25 - 2*T0*a^2*a4*b4*lambda(i) + 2*T0*a^2*a4*b4
+ T0*a*a1*b*b2 + (189*T0*a*a1*b*b3*lambda(i))/100 +
(T0*a*a1*b*b4*lambda(i))/2 - (T0*a*a1*b*b4)/2 + T0*a*a2*b*b1 +
(189*T0*a*a2*b*b3*lambda(i))/50 + T0*a*a2*b*b4*lambda(i) - T0*a*a2*b*b4 -
(T0*a*a4*b*b1)/2 - T0*a*a4*b*b2 - (189*T0*a*a4*b*b3*lambda(i))/50 -
T0*a*a4*b*b4*lambda(i) + T0*a*a4*b*b4 + (T0*a2*b^2*b2)/2 +
(189*T0*a2*b^2*b3*lambda(i))/200 + (T0*a2*b^2*b4*lambda(i))/4 -
(T0*a2*b^2*b4)/4 - (T0*a4*b^2*b2)/4 - (189*T0*a4*b^2*b3*lambda(i))/400 -
(T0*a4*b^2*b4*lambda(i))/8 + (T0*a4*b^2*b4)/8 +
(189*Ti^2*a^2*b1*b3*lambda(i))/50 + Ti^2*a^2*b1*b4*lambda(i) +
(35721*Ti^2*a^2*b3^2*lambda(i)^2)/2500 +
(189*Ti^2*a^2*b3*b4*lambda(i)^2)/25 - (189*Ti^2*a^2*b3*b4*lambda(i))/50 +
Ti^2*a^2*b4^2*lambda(i)^2 - Ti^2*a^2*b4^2*lambda(i) +
(189*Ti^2*a*b*b1*b3*lambda(i))/200 + (Ti^2*a*b*b1*b4*lambda(i))/4 +
(189*Ti^2*a*b*b2*b3*lambda(i))/100 + (Ti^2*a*b*b2*b4*lambda(i))/2 +
(35721*Ti^2*a*b*b3^2*lambda(i)^2)/5000 +
(189*Ti^2*a*b*b3*b4*lambda(i)^2)/50 - (189*Ti^2*a*b*b3*b4*lambda(i))/100 +
(Ti^2*a*b*b4^2*lambda(i)^2)/2 - (Ti^2*a*b*b4^2*lambda(i))/2 +
(189*Ti^2*b^2*b2*b3*lambda(i))/400 + (Ti^2*b^2*b2*b4*lambda(i))/8 +
(35721*Ti^2*b^2*b3^2*lambda(i)^2)/40000 +
(189*Ti^2*b^2*b3*b4*lambda(i)^2)/400 - (189*Ti^2*b^2*b3*b4*lambda(i))/800 +
(Ti^2*b^2*b4^2*lambda(i)^2)/16 - (Ti^2*b^2*b4^2*lambda(i))/16 +
(189*Ti*a^2*a3*b1*lambda(i))/25 + (35721*Ti*a^2*a3*b3*lambda(i)^2)/1250 +
(189*Ti*a^2*a3*b4*lambda(i)^2)/25 - (189*Ti*a^2*a3*b4*lambda(i))/25 +
2*Ti*a^2*a4*b1*lambda(i) + (189*Ti*a^2*a4*b3*lambda(i)^2)/25 +
2*Ti*a^2*a4*b4*lambda(i)^2 - 2*Ti*a^2*a4*b4*lambda(i) +
(189*Ti*a*a3*b*b1*lambda(i))/100 + (189*Ti*a*a3*b*b2*lambda(i))/50 +
(35721*Ti*a*a3*b*b3*lambda(i)^2)/2500 + (189*Ti*a*a3*b*b4*lambda(i)^2)/50 -
(189*Ti*a*a3*b*b4*lambda(i))/50 + (Ti*a*a4*b*b1*lambda(i))/2 +
Ti*a*a4*b*b2*lambda(i) + (189*Ti*a*a4*b*b3*lambda(i)^2)/50 +
Ti*a*a4*b*b4*lambda(i)^2 - Ti*a*a4*b*b4*lambda(i) +
(189*Ti*a3*b^2*b2*lambda(i))/200 + (35721*Ti*a3*b^2*b3*lambda(i)^2)/20000 +
(189*Ti*a3*b^2*b4*lambda(i)^2)/400 - (189*Ti*a3*b^2*b4*lambda(i))/400 +
(Ti*a4*b^2*b2*lambda(i))/4 + (189*Ti*a4*b^2*b3*lambda(i)^2)/400 +
(Ti*a4*b^2*b4*lambda(i)^2)/8 - (Ti*a4*b^2*b4*lambda(i))/8 + a^2*a1^2 +
(189*a^2*a1*a3*lambda(i))/25 + 2*a^2*a1*a4*lambda(i) - 2*a^2*a1*a4 +
(35721*a^2*a3^2*lambda(i)^2)/2500 + (189*a^2*a3*a4*lambda(i)^2)/25 -
(189*a^2*a3*a4*lambda(i))/25 + a^2*a4^2*lambda(i)^2 - 2*a^2*a4^2*lambda(i)
+ a^2*a4^2 + 24000*HCV*a^2*b1 + 90720*HCV*a^2*b3*lambda(i) +
24000*HCV*a^2*b4*lambda(i) - 24000*HCV*a^2*b4 + a*a1*a2*b +
(189*a*a1*a3*b*lambda(i))/100 + (a*a1*a4*b*lambda(i))/2 - (a*a1*a4*b)/2 +
(189*a*a2*a3*b*lambda(i))/50 + a*a2*a4*b*lambda(i) - a*a2*a4*b +
(35721*a*a3^2*b*lambda(i)^2)/5000 + (189*a*a3*a4*b*lambda(i)^2)/50 -
(189*a*a3*a4*b*lambda(i))/50 + (a*a4^2*b*lambda(i)^2)/2 -
a*a4^2*b*lambda(i) + (a*a4^2*b)/2 + 2000*HCV*a*b*b1 + 12000*HCV*a*b*b2 +
30240*HCV*a*b*b3*lambda(i) + 8000*HCV*a*b*b4*lambda(i) - 8000*HCV*a*b*b4 +
(a2^2*b^2)/4 + (189*a2*a3*b^2*lambda(i))/200 + (a2*a4*b^2*lambda(i))/4 -
(a2*a4*b^2)/4 + (35721*a3^2*b^2*lambda(i)^2)/40000 +
(189*a3*a4*b^2*lambda(i)^2)/400 - (189*a3*a4*b^2*lambda(i))/400 +
(a4^2*b^2*lambda(i)^2)/16 - (a4^2*b^2*lambda(i))/8 + (a4^2*b^2)/16 +
1000*HCV*b^2*b2 + 1890*HCV*b^2*b3*lambda(i) + 500*HCV*b^2*b4*lambda(i) -
500*HCV*b^2*b4)^(1/2) + 756*a*a3*lambda(i) + 200*a*a4*lambda(i) +
189*a3*b*lambda(i) + 50*a4*b*lambda(i))/(200*a*b1 - 200*a*b4 + 100*b*b2 -
50*b*b4 + 756*a*b3*lambda(i) + 200*a*b4*lambda(i) + 189*b*b3*lambda(i) +
50*b*b4*lambda(i));

fprintf('\nFlame temperature = %f K',T) % Adiabatic flame temperature

c = (1.013 * P0 * 10^6)/(8.314 * T);

% Rate constants for Zeldovich mechanism in m^3/mol-s


K_1 = 1.8 * 10^8 * exp( - 38370/T);
K_2 = 1.8 * 10^4 * exp( - 4680/T);
K_3 = 7.1 * 10^7 * exp( - 450/T);

% Equilibrium constants
K1 = 4.71 * exp( - 10900/T);
K2 = 3030 * exp( - 57830/T);
K3 = 3030 * exp( - 30790/T);
K4 = 166 * exp( - 19680/T);

% Equilirium concentrations
yNOe = K1 * ((yN2 * yO2)^0.5);
yNe = K2 * ((yN2/p1)^0.5);
yOe = K3 * ((yO2/p1)^0.5);
yOHe = K4 * ((yO2^0.25) * (yH2O^0.5))/(p1^0.25);

% One - way reaction Rates


R1 = K_1 * c^2 * yN2 * yOe;
R2 = K_2 * c^2 * yNe * yO2;
R3 = K_3 * c^2 * yNe * yOHe;

k = R1 / (R2 + R3);
TNO = c * yNOe / (4 * R1); % Characteristic time for NO formation

syms ('Alpha','positive');
Alpha = solve(((1 - k) * log(1 + Alpha)) - ((1 + k) * log(1 - Alpha)) -
tr/TNO,'Alpha');
Alpha = double(Alpha);
yNO = Alpha * yNOe;
RNO = 2 * R1 * (1 - Alpha^2) / (1 + k*Alpha);

fprintf('\nratio=%f
ppm\nlambda(i)=%f\nalpha=%f\nyNO=%f\nRNO=%f',tr/TNO,lambda(i),Alpha,yNO
*10^6,RNO);

subplot(2,2,1)
plot(lambda(i),T,'-b')
title('Lambda Vs Temperature');
xlabel('Lambda');
ylabel('Temperature(K))');
grid on;
hold on

subplot(2,2,2)
plot(lambda(i),yNO*10^6,'-b')
title('Lambda Vs NO concentration(ppm)');
xlabel('Lambda');
ylabel('NO concentration(ppm)');
grid on;
hold on

subplot(2,2,3)
plot(tr/TNO,Alpha,'-b')
title('tr/TNO Vs Alpha');
xlabel('tr/TNO');
ylabel('Alpha');
grid on;
hold on

subplot(2,2,4)
plot(lambda(i),RNO,'-r')
title('lambda Vs NO Rate ');
xlabel('lambda');
ylabel('NO Rate');
grid on
hold on

end

if lambda(i)<1.07
x =
[11.869 11.936 12.002 12.069 12.135 12.202 12.269 12.336 12.403
12.469 12.536 12.603 12.67 12.737 12.803 12.87 12.937 13.071
13.037 13.137 13.204 13.271 13.338 13.405 13.472 13.539 13.605
13.739 13.705 13.871 13.938 14.037 14.202 14.334 14.5 14.631
14.862];

y =
[967.515 1090.323 1201.905 1302.262 1402.62 1525.428
1648.236 1771.044 1916.303 2039.111 2150.694 2262.276
2396.31 2541.569 2653.152 2798.411 2898.768 3189.286
3066.142 3312.094 3423.677 3580.161 3714.195 3848.228
3993.487 4116.295 4239.104 4428.592 4339.125 4550.728
4662.311 4694.979 4850.455 4905.238 4982.136 5025.694
5068.243];

e = polyfit(x,y,4);
yNO = polyval(e,lambda(i)*13.88);

fprintf('\nyNO=%f\nlambda(i)=%f',yNO,lambda(i))

subplot(2,2,2)
plot(lambda(i),yNO,'-b')
title('Lambda Vs NO concentration(ppm)');
xlabel('Lambda');
ylabel('NO concentration(ppm)');
grid on;
hold on

end
fprintf('\ni=%d \n',i)
end

Output:
FORMATION OF NOx
Considering the heterogeneous nature of fuel-air mixture in diesel engines NOx and
Particulate matter (PM) are important emissions. Continuous efforts are being made to
minimize the quantities of these two pollutants from the diesel engine exhaust. Modelling and
gas sampling studies have been done with indirect injection(IDI) diesel engines, which suggest
that pre chamber is the prominent location for formation of nitrogen oxides. There are a number
of potential mechanisms responsible for NO in combustion processes. The relative importance of
these different mechanisms is strongly affected by the temperature, fuel-air equivalence ratio,
pressure, flame conditions, residence time and concentrations of key reacting species. Rapid
NOx formation begins after the start of heat release. Shortly after the end of heat release, the
period of rapid NOx formation ends because temperatures of the burned gas decrease due to
mixing with cool bulk gas and expansion of the charge. Fuel-Air equivalence ratio is another
important factor influencing NOx formation. As the equivalence ratio becomes leaner, NO and
NOx decrease significantly as expected. NO2 however shows an opposite trend to that of NO that
causes the NO2/NOx ratio to increase at leaner conditions. The NO2peaks at an equivalence ratio
near 0.25. Leaner equivalence ratio is indicative flower loads and lower bulk gas temperatures
that are conducive to the formation of NO2.
Advancing injection timing or increasing injection pressure improves combustion
efficiency raises combustion temperature. In general, higher combustion temperatures lead to
higher NOx formation. Addition of diluents to the engine intake air is considered as an
effective mean to reduce the NO formation rate and hence the exhaust NOx levels. The effect is
primarily one of reducing the peak flame temperature, which is the driving factor for NOx
formation. Diluents such as N2, CO2 and exhaust gas were added to the intake air of direct
injection (DI) engine to study their effect on NOx reduction. NOx emissions comprise of NO
and NO2. The NO2 is formed via NO molecule. Therefore, the modelling of NOx formation is
most often reduced to studying the formation of NO. It is widely accepted that in diesel engines
the major portion of NO is formed via thermal path.
Kinetics of NO Formation:-
While Nitric oxide (NO) and nitrogen dioxide (NO,) are usually grouped together as
NOx, emissions, nitric oxide is the predominant oxide of nitrogen produced inside the engine
cylinder. The principal source of NO is the oxidation of atmospheric (molecular) nitrogen.
However, if the fuel contains significant nitrogen the oxidation of the fuel nitrogen-containing
compounds is an additional source of NO. Gasoline contains negligible amounts of nitrogen;
although diesel fuel contain more nitrogen, current levels are not significant.
The mechanism of NO formation from atmospheric nitrogen has been studied
extensively.' It is generally accepted that in combustion of near stoichiometric fuel-& mixtures
the principal reactions governing the formation of NO from molecular nitrogen are,
O+N2=NO+N (i)
N + O2 = NO + O (ii)
N+OH=NO+H (iii)

Fig: - Rate constants for NO formation mechanism

The rate of formation of NO is given by,


The same reaction for rate of formation of N is given by,

The NO rate formation becomes,

Where

To introduce this equilibrium assumption it is convenient to use the notation,


R1 = K1+ [O]e [N]e
= K1-[NO]e [N]e
Where [ ] denotes equilibrium concentration for the one-way equilibrium rate.
The above eqn. reduces to,

The values of R1, R2 and R1/ (R2+R3) are given in the table.
The strong dependence of d [NO]/dt on temperature in the exponential term is evident.
High temperatures and high oxygen concentrations result in high NO formation rates.The
Figure shows the NO formation rate as a function of gas temperature and fuel/air equivalence
ratio in post flame gases. Also shown is the adiabatic flame temperature attained by a fuel-air
mixture initially at 700 K at a constant pressure of 15 atm. For adiabatic constant-pressure
combustion (an appropriate model for each element of fuel that burns in an engine), this initial
NO formation rate peaks at the stoichiometric composition, and decreases rapidly as the mixture
becomes leaner or richer.
Formation Time: - A characteristic time for the NO formation process can be defined by,
(a) NO and NO2 concentrations in SI engine as function of air/fuel ratio,
(b) NO2 as percent of total NOx in diesel exhaust as function of load and speed.

The fig. shows examples of NO and NO emissions data from a spark ignition and a diesel
engine. The maximum value for the ratio (NO2/NO) for the SI engine is 2 percent at an
equivalence ratio of about 0.85. For the diesel this ratio is higher and is highest at light load and
depends on engine speed.
NO Formation in Spark-Ignition Engines:-
In conventional spark-ignition engines the fuel and air (and any recycled exhaust) are mixed
together in the engine intake system, and vigorous mixing with the residual gas within the
cylinder occurs during the intake process. Thus the fuel/air ratio and the amount of diluents
(residual gas plus any recycled exhaust) is approximately uniform throughout the charge within
the cylinder during combustion.

Fig: - (a) Measured cylinder pressure p and calculated mass fraction burned Xb

Fig: - (b) Calculated temperature of unburned gas and burned gas


in early- and late burning elements

Fig: - (c) Calculated NO concentrations in early- and late-burning elements


For rate controlled model and at equilibrium.
The NO concentration gradient across gas in the engine cylinder, due to the temperature
gradient, has been demonstrated sampling techniques and using measurements of the
chemiluminescent radiation from the reaction,
NO +O2-- NO2.
For determining the local NO concentration with the help of figure NO concentration data as a
function of crank angle taken by Lavoie two different windows in the cylinder head of a
specially constructed L head engine where each window was a different distance from the spark
plug.

Fig: - Spectroscopic ally measured NO concentrations through two windows W3 and W2 in special L-
head SI engine (W2 is closer to spark than W3).The asterisks mark estimated initial conditions and flame
arrival times. The dashed lines are calculated rate limited concentrations for parts of charge burning at
these flame arrival times with zero initial NO concentration.
Equivalence ratio: -
Fig shows the effect of variations in the fuel air equivalence ratio on NO emissions.
Maximum burned gas temperatures occur at equivalence ratio = 1.1 however at this equivalence
ratio oxygen concentrations are low. As the mixture is enriched burned gas temperatures fall and
as the mixture is leaned out increasing oxygen concentration initially offsets the falling gas
temperatures and NO emissions peak at 0.9. Detailed predictions of NO concentrations in the
burned gases suggest that the concentration versus time histories under fuel-lean conditions is
different in character from those for fuel-rich conditions. In lean mixtures NO concentrations
freeze early in the expansion process and little NO decomposition occurs. In rich mixtures,
substantial NO decomposition occurs from the peak concentrations present when the cylinder
pressure is a maximum Thus in lean mixtures, gas conditions at the time of peak pressure are
especially significant.

Fig: - Variation of exhaust NO concentration


With A/F and fuel/air equivalence ratio.

Spark Timing: -
Spark timing significantly affects NO emission level. advancing the timing so that
combustion occurs earlier in the cycle increases the peak cylinder pressure (because more fuel is
burned before TC and the peak pressure moves closer to TC where the cylinder volume is
smaller) retarding the timing decreases the peak cylinder pressure (because more of the fuel
burns after TC). Higher peak cylinder pressures result in higher peak burned gas temperatures,
and hence higher NO formation rates. For lower peak cylinder pressures lower NO formation
rates result.
The figure shows typical NO emission data for a spark-ignition engine as a function of
spark timing. NO emission levels steadily decrease as spark timing is retarded from MBT timing
and moved closer to TC. Since exact determination of MBT timing is difficult (and not critical
for fuel consumption and power where the variation with timing around MBT is modest) there is
always considerable uncertainty in NO emissions at MBT timing.

Fig: -Variation of exhaust NO concentration with spark retard.


NOx Formation in Compression-Ignition Engines: -
The kinetic mechanisms for NO and NO, formation described in above Sections and the
assumptions made regarding equilibration of species in the C-0-H system apply to diesels as well
as to spark-ignition engines. The critical difference of course is that injection of fuel into the
cylinder occurs just before combustion starts and that non uniform burned gas temperature and
composition result from this non uniform fuel distribution during combustion. The fuel-air
mixing and combustion processes are extremely complex. During the mixing, Controlled
combustion phase the burning mixture is likely to be closer to stoichiometric (the flame structure
is that of a turbulent, though unsteady diffusion flame). However, throughout the combustion
process mixing between already burned gases, air and lean and rich unburned fuel vapour-air
mixture occurs changing the composition of any gas elements that burned at a particular
equivalence ratio. In addition to these composition (and hence temperature) changes due to
mixing, temperature changes due to compression and expansion occur as the cylinder pressure
rises and falls. The critical equivalence ratio for NO formation in high-temperature high-pressure
burned gases typical of engines is close to stoichiometric. Figure shows the initial NO formation
rate in combustion products formed by burning a mixture of a typical hydrocarbon fuel with air
(initially at 700 K at a constant pressure of 15 atm). NO formation rates are within a factor of 2
of the maximum value for 0.85< Equ. Ratio< 1.1.

Fig: - Concentrations of soot, NO, and other combustion product species measured at
outer edge of bowl in piston combustion chamber of quiescent DI diesel with rapid sampling valve.
The critical time period is when gas temperatures are at a maximum burned. Mixture
which burns early in the combustion process is especially important since it is compressed to a
higher temperature increasing the NO formation rate as combustion proceeds and cylinder
pressure increases. After the time of peak pressure burned gas temperatures decrease as the
cylinder gases expand. The decreasing temperature due to expansion and due to mixing of High-
temperature gas with air or cooler burned gas freezes the NO chemistry. This second effect
(which occurs only in the diesel) means that freezing occurs more rapidly in the diesel than in the
spark-ignition engine, and much less decomposition of the NO occurs. The above description is
supported by the NO concentration data obtained from experiments where gas was sampled from
within the cylinder of normally operating diesel engines with special gas-sampling valves and
analysed. Figure shows time histories of major species concentrations, through the combustion
process determined with a rapid-acting sampling valve (1 ms open time) in a quiescent direct-
injection diesel engine.
Equivalence Ratio:-
The effect of the overall equivalence ratio on NO concentrations is shown in the below
Fig. At high load with higher peak pressures and hence temperatures and larger regions of close-
to-stoichiometric burned gas NO levels increase. NO2 is 10 to 20 percent of total NOx. Though
NO levels decrease with a decreasing overall equivalence ratio they do so much less rapidly than
do spark-ignition engine NO emissions due to the non-uniform fuel distribution in the diesel.
Though the amount of fuel injected decreases proportionally as the overall equivalence ratio is
decreased much of the fuel still burns close to stoichiometric. Thus NO emissions should be
roughly proportional to the mass of fuel injected.

Fig: -Exhaust NOx and NO concentrations as a function of overall equivalence ratio.


NOx Prediction with change in temperature and A/F Ratio

Theory:

Background :
Principally, NOx can be formed in three different ways: prompt NOxformation,thermalNOx
formation and fuel NOx formation. In the prompt formation, the nitrogen in the air reacts with
unburned hydrocarbons ,where nitrogen radicals are produced. The free radicals are reactants in
further chemical reactions where NOx is finally formed. The fuel NOx formation can only take
place if there are nitrogen atoms chemically bound in the fuel used for combustion. When the
bounded nitrogen reacts with the oxygen in the air, NOx is produced. Although there are three
different ways for NOx to be formed, the most contributing mechanism to NOx production is the
thermal NOx formation. In this case the combustion provides sufficiently high temperatures for
the stable oxygen and nitrogen molecules to react and form NOx.A common assumption is that
the NOx formed is equal to the formed. This is however not completely true due to the fact that
there are several reaction mechanisms that will form molecules as well. Nevertheless, the
assumption is somewhat valid and practical for further modeling of NOx formation.
Considering the heterogeneous nature of fuel-air mixture in diesel
enginesNOxandParticulate matter (PM) are important emissions. Continuous efforts are
beingmadetominimize the quantities of these two pollutants from the diesel engineexhaust.
Modelling and gas sampling studies have been done with indirect injection(IDI) diesel engines,
which suggest that pre chamber is the prominent location forformation of nitrogen oxides.
There are a number of potential mechanisms responsibleforNO in combustion processes. The
relative importance of these different mechanismsis strongly affected by the temperature, fuel-
air equivalence ratio, pressure, flame conditions, residence time and concentrations of key
reacting species. Rapid NOxformation begins after the start of heat release. Shortly after the end
of heatrelease, the period of rapid NOx formation ends because temperatures of theburned gas
decrease due to mixing with cool bulk gas and expansion of the charge. Fuel-Air equivalence
ratio is another important factorinfluencingNOx formation. As the equivalence ratio becomes
leaner, NO and NOxdecrease significantly as expected. NO2 however shows an opposite trend to
thatof NO that causes the NO2/NOx ratio to increase at leaner conditions. The NO2peaks at an

1
equivalence ratio near 0.25. Leaner equivalence ratio is indicative flower loads and lower bulk
gas temperatures that are conducive to the formationof NO2.
Advancing injection timing or increasing injection pressure improves
combustionefficiency raises combustion temperature. In general, higher combustion
temperatures lead to higher NOxformation.Additionofdiluents to the engine intake air is
considered as an effective mean to reduce theNO formation rate and hence the exhaust NOx
levels. The effect is primarily oneof reducing the peak flame temperature, which is the driving
factor for NOxformation.Diluents such as N2, CO2 and exhaust gas were added to the intake air
ofdirect injection (DI) engine to study their effect on NOx reduction. NOx emissions comprise
of NO and NO2. The NO2 is formed via NO molecule.Therefore, the modelling of NOx
formation is most often reduced to studying theformation of NO. It is widely accepted that in
diesel engines the major portion ofNO is formed via thermal path.

Calculation of NOx formation rate:-


While Nitric oxide (NO) and nitrogen dioxide (NO,) are usually grouped together as
NOx, emissions, nitric oxide is the predominantoxide of nitrogen produced inside the engine
cylinder. The principal source of NO is the oxidation of atmospheric (molecular) nitrogen.
However, if the fuel contains significant nitrogen the oxidation of the fuel nitrogen-containing
compounds is an additional source of NO. Gasoline contains negligible amounts of nitrogen;
although diesel fuel contain more nitrogen, current levels are not significant.
The mechanism of NO formation from atmospheric nitrogen has been studied
extensively.' It is generally accepted that in combustion of near stoichiometric fuel-& mixtures
the principal reactions governing the formation of NO from molecular nitrogen are,
O+N2=NO+N (i)
N + O2 = NO + O (ii)
N+OH=NO+H (iii)

2
Fig: - Rate constants for NO formation mechanism

The rate of formation of NO is given by,

d [ NO]
k1 [O][ N 2 ] k 2 [ N ][O2 ] k 3 [ N ][OH ] k1 [ NO][ N ] k 2 [ NO][O] k 3 [ NO][ H ]
dt
The same reaction for rate of formation of N is given by,

d[ N ]
k1 [O][ N 2 ] k 2 [ N ][O2 ] k 3 [ N ][OH ] k1 [ NO][ N ] k 2 [ NO][O] k 3 [ NO][ H ]
dt

The NO rate formation becomes,

d [ NO] 1 [ NO]2 /( K [O2 ][ N 2 ])


2k1 [O][ N 2 ]
dt
1 k1 [ NO] / k 2 [O2 ] k 3 [OH ]
Where


K k1 / k1 k 2 / k 2

3
To introduce this equilibrium assumption it is convenient to use the notation,
R1 = K1+ [O]e [N]e
= K1-[NO]e [N]e
Where [ ] denotes equilibrium concentration for the one-way equilibrium rate.
The above eqn. reduces to,

d [ NO] 2 R1{1 ([ NO] /[ NO]e ) 2 }



dt 1 ([ NO] /[ NO]e ) R1 /( R2 R3 )
The values of R1, R2 and R1/ (R2+R3) are given in the table.

The strong dependence of NO formation rate can be demonstrated by considering the


initial value of d[NO]/dt when [NO]/[NO]e<<
Then from above eqn,
d [ NO]
2 R1 2k1 [O] e [ N 2 ] e
dt
Where,
38000
k1 7.6 1013 exp( ) cm^3/gmol.s
T
[O]e [O2]e
The can be written in the form of [O]e = K(T)
Which results in theexpression:

4
d [ NO]
2k1 k (T ) [O2] e [ N 2 ] e
dt
One lesson from above equation is that the NOx formation is dependent ontemperature, oxygen
concentrations and the concentration of nitrogen. This is anexpected result from the experimental
research conducted by Decetal.

High temperatures and high oxygen concentrations result in high NO formation


rates.TheFigure shows the NO formation rate as a function of gas temperature and fuel/air
equivalence ratio in post flame gases. Also shown is the adiabatic flame temperature attained by
a fuel-air mixture initially at 700 K at a constant pressure of 15 atm. For adiabatic constant-
pressure combustion (an appropriate model for each element of fuel that burns in an engine), this
initial NO formation rate peaks at the stoichiometric composition, and decreases rapidly as the
mixture becomes leaner or richer.

Fig.Initial NO formation rate, mass fraction per second (for [NO]/[NO]e<<1), as a function of
temperature for different equivalence ratios ( ) and 15 atm pressure. Dashed line shows
adiabatic flame temperature for kerosene combustion with 700k , 15 atm air.

5
1

PREDICTION OF NOx PROBLEM 11.10

Problem statement:
Nitric oxide, NO, forms via reactions (1) to (3), Reaction (1) is "slow" and reactions (2) and (3)
are "fast," so theInitial rate of formation of NO is given by Eq.
d [ NO]
2 R1 2k1 [O]e [ N 2 ]e
dt
Where,
38000
k1 7.6 1013 exp( ) cm^3/gmol.s
T
where [ ] denote concentrations in gram-moles per cubic centimeter, k: is the rate constant for
reaction (1), and the factor of 2 enters because the N atom formed in (1) immediately reacts via
(2) or (3) to give an additional NO molecule.
where T is in kelvin.
Using the equilibrium composition data provided for mole fraction atomicoxygen (0), molecular
nitrogen (N,), and nitric oxide (NO):

(a) Plot the formation rate of NO as a function of the equivalence ratio at 3000K and 5.5
MPa, and as a function of temperature for a stoichiometric mixture at 5.5 MPa.

(b) Estimate approximately the time taken to reach equilibrium NO levels at$ = 1, 2750 K
and 3000 K, 5.5 MPa.

(c) If the stoichiometric mixture inducted into the engine reaches 3000 K and5.5 MPa after
combustion, in the absence of any exhaust gas recirculation, calculate the percentage of
the exhaust that must be recycled to the intake (at the initial intake temperature) to
reduce the NO formation rate by a factor of 4 (assume the final pressure 5.5 MPa stays
the same; of course, the final temperature decreases as the exhaust gas is recycled).
2

Theory:

FORMATION OF NOx:
Considering the heterogeneous nature of fuel-air mixture in diesel
enginesNOxandParticulate matter (PM) are important emissions. Continuous efforts are
beingmade tominimize the quantities of these two pollutants from the diesel engineexhaust.
Modelling and gas sampling studies have been done with indirect injection(IDI) diesel engines,
which suggest that pre chamber is the prominent location forformation of nitrogen oxides.
There are a number of potential mechanisms responsibleforNO in combustion processes. The
relative importance of these different mechanismsis strongly affected by the temperature, fuel-
air equivalence ratio, pressure, flame conditions, residence time and concentrations of key
reacting species. Rapid NOxformation begins after the start of heat release. Shortly after the end
of heatrelease, the period of rapid NOx formation ends because temperatures of theburned gas
decrease due to mixing with cool bulk gas and expansion of the charge. Fuel-Air equivalence
ratio is another important factorinfluencing NOx formation. As the equivalence ratio becomes
leaner, NO and NOxdecrease significantly as expected. NO2 however shows an opposite trend to
thatof NO that causes the NO2/NOx ratio to increase at leaner conditions. The NO2peaks at an
equivalence ratio near 0.25. Leaner equivalence ratio is indicative flower loads and lower bulk
gas temperatures that are conducive to the formationof NO2.
Advancing injection timing or increasing injection pressure improves
combustionefficiency raises combustion temperature. In general, higher combustion
temperatures lead to higher NOx formation.Addition ofdiluents to the engine intake air is
3

considered as an effective mean to reduce theNO formation rate and hence the exhaust NOx
levels. The effect is primarily oneof reducing the peak flame temperature, which is the driving
factor for NOx formation.Diluents such as N2, CO2 and exhaust gas were added to the intake air
ofdirect injection (DI) engine to study their effect on NOx reduction. NOx emissions comprise
of NO and NO2. The NO2 is formed via NO molecule.Therefore, the modelling of NOx
formation is most often reduced to studying theformation of NO. It is widely accepted that in
diesel engines the major portion ofNO is formed via thermal path.

Kinetics of NO Formation:-
While Nitric oxide (NO) and nitrogen dioxide (NO,) are usually grouped together as
NOx, emissions, nitric oxide is the predominantoxide of nitrogen produced inside the engine
cylinder. The principal source of NO is the oxidation of atmospheric (molecular) nitrogen.
However, if the fuel contains significant nitrogen the oxidation of the fuel nitrogen-containing
compounds is an additional source of NO. Gasoline contains negligible amounts of nitrogen;
although diesel fuel contain more nitrogen, current levels are not significant.
The mechanism of NO formation from atmospheric nitrogen has been studied
extensively.' It is generally accepted that in combustion of near stoichiometric fuel-& mixtures
the principal reactions governing the formation of NO from molecular nitrogen are,
O+N2=NO+N (1)
N + O2 = NO + O (2)
N+OH=NO+H (3)
4

Fig: - Rate constants for NO formation mechanism

The rate of formation of NO is given by,

d [ NO]
k1 [O][ N 2 ] k 2 [ N ][O2 ] k 3 [ N ][OH ] k1 [ NO][ N ] k 2 [ NO][O] k 3 [ NO][ H ]
dt
The same reaction for rate of formation of N is given by,

d[ N ]
k1 [O][ N 2 ] k 2 [ N ][O2 ] k 3 [ N ][OH ] k1 [ NO][ N ] k 2 [ NO][O] k 3 [ NO][ H ]
dt

The NO rate formation becomes,

d [ NO] 1 [ NO]2 /( K [O2 ][ N 2 ])


2k1 [O][ N 2 ]
dt
1 k1 [ NO] / k 2 [O2 ] k 3 [OH ]
Where


K k1 / k1 k 2 / k 2
5

To introduce this equilibrium assumption it is convenient to use the notation,


R1 = K1+ [O]e [N]e
= K1-[NO]e [N]e
Where [ ] denotes equilibrium concentration for the one-way equilibrium rate.
The above eqn. reduces to,

d [ NO] 2 R1{1 ([ NO] /[ NO]e ) 2 }



dt 1 ([ NO] /[ NO]e ) R1 /( R2 R3 )
The values of R1, R2 and R1/ (R2+R3) are given in the table.

The strong dependence of NO formation rate can be demonstrated by considering the


initial value of d[NO]/dt when [NO]/[NO]e<<
Then from above eqn,
d [ NO]
2 R1 2k1 [O]e [ N 2 ]e
dt
Where,
38000
k1 7.6 1013 exp( ) cm^3/gmol.s
T
High temperatures and high oxygen concentrations result in high NO formation
rates.TheFigure shows the NO formation rate as a function of gas temperature and fuel/air
equivalence ratio in post flame gases. Also shown is the adiabatic flame temperature attained by
a fuel-air mixture initially at 700 K at a constant pressure of 15 atm. For adiabatic constant-
6

pressure combustion (an appropriate model for each element of fuel that burns in an engine), this
initial NO formation rate peaks at the stoichiometric composition, and decreases rapidly as the
mixture becomes leaner or richer.

Fig.Initial NO formation rate, mass fraction per second (for [NO]/[NO]e<<1), as a function of
temperature for different equivalence ratios ( ) and 15 atm pressure. Dashed line shows
adiabatic flame temperature for kerosene combustion with 700k , 15 atm air.

Formation Time: - Acharacteristic time for the NO formation process can be defined by,

1 d [ NO]
NO
1

[ NO] e dt

8 10 16 T exp(58,300 / T )
NO
p1 / 2

Where NO is in seconds, T in kelvins, and p in atm. Use has been made of the fact that

Xn2=0.71 for engine combustion conditions,


NO is usually to or longer than the times
7

characteristics of changes in engine conditions so the formation process is kinetically controlled.


However for close stoichiometric conditions at the maximum pressure and burned gas

temperatures NO is of the same order as typical combustion times (1ms) and equilibrium
concentration may be attained
8

(a) NO and NO2 concentrations in SI engine as function of air/fuel ratio,

Equivalence ratio: -
Fig shows the effect of variations in the fuel air equivalence ratio on NO emissions.
Maximum burned gas temperatures occur at equivalence ratio = 1.1 however at this equivalence
ratio oxygen concentrations are low. As themixture is enriched burned gas temperatures fall and
as the mixture is leaned out increasing oxygen concentration initially offsets the falling gas
temperatures andNO emissions peak at 0.9. Detailed predictions of NO concentrations in
theburned gases suggest that the concentration versus time histories under fuel-lean conditions is
different in character from those for fuel-rich conditions. In lean mixturesNO concentrations
freeze early in the expansion process and little NOdecomposition occurs. In rich mixtures,
substantial NO decomposition occurs from the peak concentrations present when the cylinder
pressure is a maximum Thus in lean mixtures, gas conditions at the time of peak pressure are
especially significant.

Fig: - Variation of exhaust NO concentration


With A/F and fuel/air equivalence ratio.
9

Fig. Effect of Oxygen concentration on NOx formation

Solution:
Inputs:
10

Matlab Program

(a)
% Plot of Formaion of NoxVs Equivalence ratio at constant
temperature 3000k
clc
N2=0.73; % concentration of N2
T=3000; % Constant temp
phi=[0.9,1,1.1]; % values of Equivalence ratio
O=[2.1e-3,1.5e-3,1e-3]; % Corresponding concentration values of
Oxygen Content
k1=7.6*(10^13)*exp(-38000/T); % Value of k1 varies with T

dNO_dt=2*k1*N2*O % NOx formation rate


plot(phi,dNO_dt)
xlabel('Equivalence Ratio (phi)');
ylabel('dNO_dt');

Output:

dNO_dt =
11

1.0e+005 *

7.3506 5.2504 3.5003

Graph:
12

% Plot of Rate of formation of NoxVs Temp at constant


Equivalence ratio (phi)=1

clc
N2=0.73; % Concentration of Nitrogen
phi=1; % Equivalence ratio
T=[2500,2750,3000]; % Range of temp
O=[6e-5,5e-4,1.5e-3]; % Range of oxygen corresponding to
temp
k=7.6*(10^13)*exp(-38000./T); % Value of k1
dNO_dt=2.*k*N2.*O % NOx formation rate
plot(T,dNO_dt)
xlabel('Temperature (degree) ');
ylabel('dNO_dt');

Output:

dNO_dt =

1.0e+005 *

0.0167 0.5533 5.2504


13

Graph:
14

(b)

% Time to reach equilibrium levels at phi=1,T=2750k,3000k


T=[2750,3000]; % Temp range
patm=1.013; % Atmospheric pressure
Time=(8e-16.*T.*exp(58000./T))/(patm^(1/2))

Ans:

Time =

0.0032 0.0006

(c)
% Percentage of exhaust gas recycle to intake to reduce Nox formation rate by factor 4
dNO_dt=max(dNO_dt);
dNO_dt=(dNO_dt/4)
dNO_dt=0.25dNO_dt
15

Ans: from graph at 25% mass reduction 5 % diluent in intake mixture needs to be recycled

References

1. John B Heywood,Internal combustion engine fundamental.


2. Simulating combustion by Guntur.
3.G. stiech, Modeling engine spray & combustion process.

You might also like