National Conference - NCAAT 2010

Advances in Automotive Technology (Proceedings of the National Conference on Advances in Automotive Technology)

Rajalakshmi Engineering College, Thandalam, Chennai, India

15th & 16th July 2010

EDITORS

Dr. S. SAMPATH

T. ASHOKKUMAR

M. RAJESH

K. MOHANRAJ Rajalakshmi Engineering College

Thandalam, Chennai

Department of Automobile Engineering

RAJALAKSHMI ENGINEERING COLLEGE

Rajalakshmi Nagar, Thandalam,

Chennai - 602 105

Department Of Automobile Engineering

RAJALAKSHMI ENGINEERING COLLEGE

Rajalakshmi Nagar, Thandalam,

Chennai - 602 105

© 2010, Rajalakshmi Engineering College, Thandalam, Chennai

All rights reserved. No parts of this publication may be reproduced or transmitted in any

form or by any means, electronically or mechanically, including photocopying, recording or

any information storage or retrieval system, without either prior permission in writing from

the publisher or a licence permitting restricted copying.

Whilst the articles in this book is published after receiving written copyright from the

authors,, believing it is original work, the publisher shall not be held responsible for any

copyright violation, if any, done by the author/s.

Published by : G.K.Publisher (Dakshin),

No.16, M.R.M. Road, C.A.R. Complex,

West Tambaram, Chennai – 600 045.

9444881098, 044-22266617

Printed by : Bhaghavan Printers,

No.16, M.R.M. Road, C.A.R. Complex,

West Tambaram, Chennai – 600 045.

9444881098, 044-22266617

Members of the Advisory Committee

Dr. V. GANESAN

Professor, IIT Chennai

Dr. A. RAMESH


Dr. G. DEVARADJANE

Professor, MIT, Anna University

Dr. A. RAJADURAI

Professor, MIT, Anna University

Dr. G. NAGARAJAN

Professor, CEG Anna University

Mr. K. VIJAYAN

Deputy Director, CTDT, Anna

University, Chennai

Dr. S. ARUMUGAM

Advisor, PET Engineering College

Dr. R. MAHADEVAN

President, India Piston Ltd.

Dr. N. RAVICHANDRAN

CEO, Lucas TVS Ltd.

Convener: Dr. S. SAMPATH

Co Convener: T. ASHOKKUMAR

Organising Secretary: M. RAJESH

Members of the Organizing Committee

Mr. A. RAMAMOORTHY

Mr. G. RAJA

Mr. R. ANBALAGAN

Mr. A. J. D. NANTHAKUMAR

Mr. P. N. SELVARAJU

Mr. K. MOHANRAJ

Mr. KIRAN VISWANATHAN

Mr B. JAYAPAL

Mr. P. BALAJI

Mr. KALIDAS Mr. BALAMURUGAN Mr. K. GOVINDARAJAN Mr. K. DAKSHINAMURTHY

Mrs. N. MEENA

Reviewers of NCAAT 2010

The editors gratefully acknowledge the valuable help rendered by the

reviewers.

Dr. V. Ganesan


Dr. G. Devaradjane

Professor, MIT,

Anna University, Chennai

Mr. A. Ramamoorthy

Asst Professor, REC,

Thandalam,, Chennai

Mr. K. Kamalakannan

Associate Professor, Hindustan

University,

Kelambakkam, Chennai

Dr. A. Ramesh


Mr. D. Moses Raja Cecil

Senior Section Engineer (Design)

ICF, Chennai

Mr. G. Raja

Asst Professor, REC,

Thandalam, Chennai.

Mr. A. J. D. Nanthakumar

Sr. Lecturer, REC,

Thandalam, Chennai

National Conference on Advances in Automotive Technology [NCAAT 2010] Rajalakshmi Engineering College, Thandalam, Chennai: 15th & 16th July 2010

MESSAGE FROM THE CHAIRPERSON

Dr. Thangam Meghanathan [M.A., M.Phil., Ph.D]

Automobile Engineering is a well established field offering excellent

opportunities to the young professionals pursuing the programme. With so

many automotive industries in India, the National Conference on NCAAT 2010

being conducted by the Department of Automobile Engineering of our college

on 15th and 16th of July 2010 assumes great significance. I understand that the

conference includes paper presentations on various facets of the branch and

also invited lectures by eminent experts from industry and academic

institutions. The Department of Automobile Engineering of the college is well

known for its positive approach in equipping the students with best industry

exposure and practices. I congratulate the team for conducting such a useful

conference and wish them a grand success.


MESSAGE FROM THE PRINCIPAL

Dr. Ranganarayanan [B.E., MSc. Engg., Ph.D]

I am very happy that the Department of Automobile Engineering of our College

is conducting a National Conference on ‘Advances in Automotive Technology

NCAAT-2010 on 15th and 16th July 2010. It is heartening to know that the

response for the Conference from institutes and industries is very good. I am

sure the conference will bring a fruitful interaction between the industries and

educational institutions.

I Congratulate the Department of Automobile Engineering for their efforts and

wish them all success.


MESSAGE FROM THE CONVENER

Dr. S. SAMPATH [B.E., M.Sc. (Engg), PhD]

H.O.D [Auto], Director [Research]

The National Conference on ‘Advances in Automotive Technology NCAAT-

2010’ aims at bringing the industries and institutes together to exchange their

ideas and work leading to further growth in their field. I am happy that the

response, from industries and engineering colleges, to the conference is very

good and 24 research papers are going to be presented. I understand that

lectures by eminent experts from industries and institutes have also been

arranged so that the participants will be exposed to the recent advances in

Automotive Technology.

I congratulate the staff of the Department of Automobile Engineering for all

their efforts in conducting this conference and wish them all success.


Preface

Considering recent developments in Automotive Technology, it becomes the need of the time for

the industries and academicians to meet and share knowledge and ideas. People involved in

education and engineering practice need to keep them updated with new concepts and tools

emerged. Keeping this in mind, the Department of Automobile Engineering of Rajalakshmi

Engineering College, decided to conduct its first National Conference on Advances in Automotive

Technology [NCAAT- 2010] on 15th and 16th July 2010.

The main objective of this conference [NCAAT 2010] is to facilitate interaction between

academicians, PG students, researchers, designers and manufacturers regarding advances in the

automotive industry. It also helps the academicians to impart knowledge of latest trends and

technologies to their students and shows the way for teaching and research in future.

The conference is attended by more than 40 participants from different states. Out of 50 papers

submitted, after rigorous peer review, 32 papers were accepted or provisionally accepted and these

papers are included in the Proceedings. Out of these, 24 papers are finally selected for oral

presentations. In addition to these, 4 plenary papers were presented by eminent researchers. The

compilation of the proceedings is categorized into the following areas.

1. Alternate fuels and emission control 2. Advances in IC Engine design 3. Design of automotive systems 4. Computational design of automotive components.

We would like to extend our utmost gratitude to the advisory committee members for their advice

and the referees for reviewing the technical contents of the articles submitted by various authors.

We wish to thank all the faculty members of the Department of Automobile Engineering of the

college for their help in editing the proceedings. We thank the supporting staff of the department

for their whole hearted involvement in the works related to NCAAT 2010. We are indebted to the

administrative staff, facilities providers and those who were involved, directly and indirectly, in the

conference related works.

We take this opportunity to thank the Management of the college for the support extended, without

which the conference wouldn’t have been possible.

Dr. S. SAMPATH

T. ASHOKKUMAR

M. RAJESH

K. MOHANRAJ


CONTENTS

PAPER

NUMBER

TITLE PAGE NO.

N-2010-D01-01 VVT Control for a Single Cylinder SI Engine 1

N-2010-D09-05 Evaluation and Simulation of Semi Active Suspension

System for Multi-Utility Cars by Modified Skyhook

Control Theory Using Full Car Model

1

N-2010-D07-08 Analysis of Ackerman Mechanism 2

N-2010-E01-09 Experimental Study on the Performance and Emission

Characteristics Using Diesel Fuel Additives

2

N-2010-E06-12 Characterization And Optimization Of Nerium Oil For

Diesel Engines

3

N-2010-F02-13 Performance Characteristics of Ethanol as an Alternate

Fuel

4

N-2010-D11-15 A Dynamic Model of a Condenser in an Automotive

Air Conditioning System

4

N-2010-S03-20 Simulation and Modelling of Evaporator in an

Automotive Air Conditioning System

5

N-2010-F01-25 Performance Evaluation of Di Diesel Engine Fuelled

With Turpentine Diesel Blends

5

N-2010-D10-26 Ergonomically designed driving system for two

wheeler

6

N-2010-D06-27 Numerical simulation using CFD and experimental

evaluation of the heat transfer rate using ethylene

glycol mixture as engine coolant

6

N-2010-S07-31 Automatic Tyre Pressure Monitoring and Control

System

7

N-2010-D02-02 Optimized Regenerative Braking System in Electric

Bike

8


N-2010-D03-03 Frontal Impact Analysis on a Heavy Passenger Vehicle 16

N-2010-D04-04 Finite Element Analysis of Inlet Manifold of an IC

Engine

24

N-2010-D08-10 Roof Mounted Driver‘s Seat for Armoured Vehicle 31

N-2010-D05-11 Parametric Study Analysis and Design Modifications of

Rear Axle Housing Assembly for Heavy Commercial

Vehicles

35

N-2010-D14-14 Piston Ring Pack Optimization for a DI Diesel Engine

by Predictive Technique Approach and Experimental

Verification

43

N-2010-E04-16 Effect of EGR and DPF on Emission of DI Diesel

Engines to Meet BS IV Norms

50

N-2010-D15-17 Utilization of vehicle frontal pressure to improve

engine‘s volumetric efficiency

57

N-2010-E01-18 Experimental Investigation on the performance and

emission of Di Diesel Engine using Eucalyptus and

Diesel in dual fuel mode

61

N-2010-S04-21 Suitability of Multi - Core for Embedded Automotive

System

66

N-2010-E08-22 GANESH for SI Engine Simulations – GUI Approach 75

N-2010-S01-23 Application of Statistical Tools in Evaluation of

Combustion Quality of SI Engines under Idling

84

N-2010-E05-24 Experimental investigation on PCCI combustion in a

single Cylinder DI engine

90

N-2010-D13-28 Experimental Study on Disc Brake Squeal 95

N-2010-S06-30 Effect of EGR on Emission Characteristics of Crude

Rice Bran Oil Blend as a CI Engine Fuel at Higher

Injection Pressure

101

N-2010-E07-33 Control of regulated emissions using intelligent

cylinder deactivation

105


N-2010-D01-01

VVT Control for a Single Cylinder SI Engine

Syed Mudhasir 1, K. R. Anandakumaran Nair

2, G. Devaradjane

3

1, 3 Madras Institute of Technology,

2 Lucas-TVS

[email protected]

ABSTRACT

Reducing fuel consumption and pollutant emissions is an imperative and a continuous challenge faced

by the automotive industry. Among those, the variable valve timing and actuation systems are of particular

interest. A new type engine valve control system has been presented, which focuses on the effects of varying the

valve timing, keeping the duration constant, on the performance and emission characteristics for a 4 stroke

single cylinder TVS-SUZUKI FIERO S.I. Engine.

The Valve Timing mechanism is made of set of planetary gears. The outer gear is the timing pulley

which has timing belt driven by the crankshaft of an engine. Three planetary gears are inside of the pulley. The

gears engage with the inner gear of the pulley, i.e. annulus. The centre of the disc, which has centres of the

planetary gear, is connected to the camshaft. Then, the crank rotation is transmitted to the camshaft, and

rotations of sun gear are added to the rotations of camshaft. This means that when rotation angle of the sun gear

is controlled, the phase between the inlet valve and the exhaust valve can be controlled. Experiments have been

performed to evaluate the effects on the performance characteristics by advancing or retarding the valve timing

based on speed and load conditions.

N-2010-D09-05

Evaluation and Simulation of Semi Active Suspension System for Multi-Utility

Cars by Modified Skyhook Control Theory Using Full Car Model

Suketu Y Jani, A.B.Mistry

L.D.C.E-A‘bad

[email protected], [email protected]

ABSTRACT

In this work, modified control theory is developed for semi active control of suspension system for MU

(Multi Utility) cars. A mathematical model for full car is developed for passive suspension system. Modified

control theory for control of damping coefficient for variable damper is included. Full car model is developed on

MATLAB/SIMULINK for results comparison for passive and semi active controlled suspension systems.

Response of Passive Suspension System and Semi Active Suspension System for defined road test profile is

compared, from which it is concluded that Semi Active Suspension System gives best results. It can be tested

for different types of road profile also. Maximum acceleration goes in Passive Suspension System is half that in

Semi Active Suspension system which comes into range of Shock Tolerance of Human Body for Comfort.

1


N-2010-D07-08

Analysis of Ackerman Mechanism

R. Venkatachalam, A. PadmaRao

National Institute of Technology, Warangal 506 004

[email protected]

ABSTRACT

Ackerman steering mechanism, though cannot always provide correct steering conditions, is very

popular because of its simple construction. A correct steering avoids skidding of the tyres, and thereby enhances

their lives as the wear of the tyres is reduced. In this paper, it is intended to analyze Ackerman mechanism and

propose a method of estimating skidding due to improper steering. Two parameters were identified using which

the length of skidding may be estimated.

N-2010-E01-09

Experimental Study on the Performance and Emission Characteristics Using

Diesel Fuel Additives

S. Kuganathan, Dr. G. Sankara Narayanan, Mr. M. Kannan

Adhi Parasakthi Engineering College, Chennai

[email protected]

ABSTRACT

The demand for compression ignition engines is continuously growing due to their good fuel efficiency.

But they cause lot of concern with regard to exhaust emissions. This concern sought to examine ways by which

the composition of the fuel used by CI engines could be changed to reduce emissions.

The addition of oxygen containing compounds to diesel fuel has been proposed as a method to complete

the oxidation of carbonaceous particulate matter and associated hydrocarbons. In addition many oxygenate have

high cetane number and their association with diesel results in high cetane number and hence lower exhaust

emissions. Due to these advantages, there is growing interest in the introduction of oxygenates into diesel fuel.

The performance and emission characteristics of two kinds of additives 2-Ethoxy Ethyl Acetate, Di-

Ethyl Ether with three different blends were investigated. A considerable reduction of Smoke emission, Carbon

monoxide and Unburned Hydrocarbon is obtained and Nitrogen Oxides emissions are increased when the oxygen

content is increased from 5% to 15%. In addition, a slight increase of brake specific fuel consumption is observed

due to the small decrease of fuel heating value with the increase of the oxygen content. Brake thermal efficiency

increased when the oxygen content was increased.

2


N-2010-E06-12

Characterization and Optimization of Nerium Oil for Diesel Engines

S. Prabhakar, V.N Banugopan, K. Annamalai, G. Devaradjane, S. Jayaraj,

P.Sentilkumar

MIT Campus, Anna University-Chennai

Email: [email protected]

ABSTRACT

The automobile sector which is growing day to day consumes the fossil fuel more than its growth. So

there is a demand for exploring new sources of fuels for existing engines. This led to the growth in bio diesel

which is an alternate fuel. An alternative fuel must be technically feasible, economically competitive,

environmentally acceptable, and readily available.

In this project esterified Nerium oil is used as an alternate fuel. A single cylinder stationary Kirloskar

engine is used to compare the performance and emission characteristics between pure diesel and Nerium blends.

In this project selection of suitable Nerium blend and selection of optimized injection timing for the blend is

done. The Nerium oil blends are in percentage of 20%, 40%, 60%, 80%, and 100% of Nerium oil to 80%, 60%,

40%, 20% and 0% of diesel.

From this project it is concluded that among all Nerium and diesel blends 20% of Nerium and 80% of

diesel blend at 30º BTDC gives better performance nearing that of diesel. When comparing the emission

characteristics HC, CO are reduced when compared to diesel, however NOx emission is slightly increased.

Hence Nerium blend can be used in existing diesel engines with minimum modification in the engine.

It also describes the usage of non-edible oil to a greater extent.

3


N-2010-F02-13

Performance Characteristics of Ethanol as an Alternate Fuel

Venkata Ramesh Mamilla, Dr.G.Lakshmi Narayana Rao, Dr.K.Rajagopal ,

V.Venkatesh


ABSTRACT

The objective of this paper is to study the performance characteristics of emulsified ethanol and

compare them with diesel. This paper reviews about different aspects of ethanol as an alternative fuel for diesel

engines. This paper discusses about properties of ethanol & compares them with diesel. . The tests were carried

out on single cylinder air cooled direct injection diesel engine.

N-2010-D11-15

A Dynamic Model of a Condenser in an Automotive Air Conditioning System

M. Arunkumar, S. Arul selvan, N. Muthukumar

Madras Institute of Technology, Chennai

ABSTRACT

This paper describes the need for dynamic (transient) simulation of automotive air conditioning

systems, particularly a component model i.e. condenser. In this project, a dynamic computer simulation of air-

conditioning condensers, based on fundamental principles, was developed. It consists of dividing the total

condenser length into few segments which are further divided into several nodes, as in the tube-by-tube

approach. Air and refrigerant heat transfer coefficients, as well as refrigerant pressure drop, were calculated

using existing correlations.

The model provides increased flexibility in terms of increased mass flow rate and refrigerant type. The

simulation is carried out in MATLAB/SIMULINK. An experimental test matrix covering a wide range of

conditions was used to validate the simulation. The model assumes that the condenser can be divided into three

distinct zones on the refrigerant side: the vapour de-superheating zone, the two-phase zone and the sub-cooled

liquid zone.

The model inputs are the air supply temperature, the air mass flow rate, the refrigerant supply

temperature (or the over-heating), the exhaust sub-cooling and the refrigerant mass flow rate. The model is able

to identify the pressures and temperatures in each zone and the corresponding heat flows. The model also gives

the geometrical repartition among the zones and the pressure drop on air-side.

4


N-2010-S03-20

Simulation and Modelling of Evaporator in an Automotive Air Conditioning System

N.Muthukumar, S. Arul selvan, M.Arunkumar


[email protected]

ABSTRACT

The focus of this project is, the development of a Mathematical Model for evaporator and Simulation is

to be performed using MATLAB SIMULINK for analysis. The modelling methodology is developed with the

multiple objectives of prediction, control and design. Firstly, the individual component, i.e. Evaporator, for a

typical subcritical cycle has been developed based on the best published theoretical and empirical literature. The

component models developed in this project were used to simulate the system response to varying component

parameters.

The evaporator model consists of two characteristics zones, evaporation and superheating. In

evaporation the air quality plays a major role and its range from inlet quality. The humidity also plays a major

role in the development of evaporator. These two are considered as performance parameters which would affect

the overall performance of the evaporator as well as that of the air conditioning system.

N-2010-F01-25

Performance Evaluation of Di Diesel Engine Fuelled With Turpentine Diesel Blends

E. Ganesh Kumar, V. Nadana Kumar, R. Karthikeyan

Adhi Parasakthi Engineering College

[email protected]

ABSTRACT

This paper presents the results of experimental work carried out to evaluate the performance

characteristics of turpentine oil fuel (TPOF) blended with conventional diesel fuel (DF) fuelled in a DI diesel

engine. Turpentine oil derived from pyrolysis mechanism of resin obtained from pine tree dissolved in a volatile

liquid can be used as a bio-fuel due to its properties. The test engine was fully instrumented to provide all the

required measurements for determination of the needed performance variables. The physical and chemical

properties of the test fuels were earlier determined in accordance to the ASTM standards. The tests indicate that

the engine operating on turpentine oil fuel at manufacture‘s injection pressure – time setting (20.5 MPa and 23

BTDC) had lower carbon monoxide (CO), unburned hydrocarbons (HC), oxides of nitrogen (NOx), smoke level

and particulate matter. Further the results showed that the addition of 30% TPOF with DF produced higher

brake power and net heat release rate with a net reduction in exhaust emissions such as CO, HC, NOx, smoke

and particulate matter. Above 30% TPOF blends, such as 40% and 50% TPOF blends, developed lower brake

power and net heat release rate due to their lower calorific value; nevertheless, reduced emissions were still

observed.

5


N-2010-D10-26

Ergonomically designed driving system for two wheeler

V. Deepan, S. Chandrasekaran, P. Arjunraj


ABSTRACT

This project is basically a combination of acceleration and braking system for two wheelers. Pedal

operated acceleration and braking system is a method in which acceleration and brakes of a two wheeler are

actuated using lever. The main purpose of installing combined braking system and acceleration is to.

Reduce wrist injury

Reduces the Human Fatigue

Easy operation of acceleration

Reduce the design complexity

Easy to travel for long distance.

The proposed new design of pedal Acceleration and Braking system is a modular type in which the

shaft is integrated with brake pedal. This system consists of two separate levers for acceleration and brake and

combined clutch lever with gear lever. This system can be employed in any two wheelers.

N-2010-D06-27

Numerical simulation using CFD and experimental evaluation of the heat

transfer rate using ethylene glycol mixture as engine coolant

Serralathan. S. R, Arunprasad. S

MIT Chennai, Anna University-Chennai

ABSTRACT

The rate of heat rejection to the coolant must be reduced to increase the performance of the engine as

the coolant side energy loss is about 33%.This heat rejection rate to the coolant depends on the coolant flow

velocity into the cylinder water jacket. The other parameter which greatly affects the coolant side heat transfer is

the coolant mixture. Normally water is used as coolant, but the heat carrying capacity of water is higher in the

nucleate boiling region. So it is necessary to go for coolant which has less heat carrying capacity during the

nucleate boiling condition. Ethylene Glycol solution has less heat carrying capacity than water at its saturation

temperature. In this work a numerical model is developed by considering the forced convection and nucleate

boiling condition. With this developed model and by varying the coolant flow velocity the reduced heat transfer

rate can be evaluated using CFD and compared with the experimental result.

6


N-2010-S07-31

Automatic Tyre Pressure Monitoring and Control System

P.K.Shyamshankar, Dr. N.S. Parthasarathy

Anna University Chennai,


ABSTRACT

This paper on ―Automatic Tyre Pressure Monitoring and Control System‖ puts forth a methodology

that facilitates the design and development of a new product that could continuously monitor and control the

tyre pressure in the vehicle when the vehicle is in motion.

Vehicles often come across different road surfaces during a travel. In order to optimize the mobility of

the vehicle, different tyre pressures are required for different types of terrain (sand, mud, cross road, tar road

etc). The tyre pressure control system will enable the vehicle operator to change the tyre pressure without

leaving the vehicle. Such a system will increase the life time of the tyre and reduce the fuel consumption. The

main purpose of the invention is to provide a reliable, economical and energy efficient means of monitoring the

tyre pressure and inflate air to the tyres of the vehicles when the vehicle is in motion. The project will provide

the vehicle, a convenient means to regulate the pressure of vehicle tyres for safe performance. The focus will be

on the design of a tyre pressure monitoring and control system for a four wheeler passenger vehicle and the

fabrication of a prototype working model for a single wheel.

7


N-2010-D02-02

Optimized Regenerative Braking System in Electric Bike

Sridhar.K1, Dr. J.Jancirani

2

Keywords: Regenerative braking, electric bike, electric brake, waste energy recovery,

ABSTRACT

Regenerative braking is an effective approach for

electric vehicles to extend their efficiency. It is the

emerging technology used on hybrid gas/electric

automobiles to recoup some of the energy lost

during stopping. Regenerative braking has to be

carried out together with the conventional barking.

In brake system design for EVs, the basic equation

must be concerned with proper application of

braking force to quickly reduce the vehicle speed

and meanwhile maintain the vehicle travelling

direction stable and controllable through the

steering wheel on various road conditions and also

recovering the braking energy as much as possible

in order to improve the energy utilization

efficiently. The regenerated energy is saved in a

storage battery and used later to power the

motor. Regenerative braking takes energy

normally wasted during braking and turns it

into usable energy. It does improve energy

efficiency of the vehicle. In this work, a

mathematical model of a regenerative braking

system for the braking efficiency has been

developed. The experimental results are compared

with the simulated results. The electricity generated

by the battery during braking varies according to

the speed of the vehicle. So, to utilize the generated

electricity completely, a suitable Electronic Control

Unit (ECU) is designed.

INTRODUCTION

A Brake System slows and stops an automobile.

Brakes are applied on the wheels to decelerate the

vehicle. In Electric vehicles, D.C. Motor is used to

propel the vehicle which draws current from

battery. Conventional braking system uses

frictional brakes to stop the vehicle. It wastes the

kinetic energy produced by the motor as heat

energy. Battery electric propulsion presents

opportunities to recover vehicle kinetic energy to

improve energy economy [1] [3]

. Regenerative

braking is used on electric automobiles to recoup

some of the energy lost during stopping. This

energy is saved in a storage battery and used later

to power the motor.

REGENERATIVE BRAKING SYSTEM

In Regenerative Braking system, as the driver

applies the brakes through a conventional pedal,

the power supply to the electric motors is stopped.

By this action, the motor stops propelling the

vehicle but due to the forward momentum the

vehicle moves for some distance and then

eventually stops the vehicle. Regenerative braking

does more than simply stop the bike.

Whenever the electric motor shaft is rotated

(mechanical energy is given to the motor), it

becomes an electric generator or dynamo. This

generated electricity is fed into a chemical storage

battery and used later to power the car at city

speeds. Regenerative braking takes energy

normally wasted during braking and turns it into

usable energy. Thus it improves energy efficiency

of the electric vehicle.

DESIGN ASPECT

The objective of this work is to implement the

Regenerative Braking technology in electric

vehicles particularly in Electric bikes. The first

phase of the work is to develop a mathematical

model and to simulate it using

MATLAB/SIMULINK.

In the second phase of the work, Regenerative

Braking System is to be implemented the in bike.

8

http://www.wisegeek.com/what-is-a-dynamo.htm

http://www.wisegeek.com/what-is-a-chemical.htm


An ECU has been developed for optimization of

the battery charging.

Fig.1.Regenerative braking

concept

EXPERIMENTAL SETUP - The experimental

setup designed for the demonstration of

regenerative braking in Electric bike is shown

below.

Fig.2.Experimental setup for RBS

Fig.3.Schematic representation of experimental

setup

Fig.4. Flange that connects motor shaft and

driveline system

Fig.5. Contact breaker in driving and charging

position

Electric Vehicle Conversion

1. A vehicle that is light and aerodynamic in

order to maximize distance travelled per

battery charge is selected. There must be also

an adequate room to load motor and batteries.

2. The battery pack, which provides a source of

electrical power. The most commonly

available and affordable batteries are lead acid

flooded type.

3. The D.C motor that propels the vehicle is

mounted on the bike chassis as shown in

figure2.

4. Electric motor shaft (Driving shaft) is

mechanically attached to the driveline system

by a flange specially designed for this purpose

(shown in figure 4).

5. The power controller, which regulates the flow

of energy between the battery and the electric

motor is controlled by an electronic throttle.

9

http://en.wikipedia.org/wiki/Battery_pack

http://en.wikipedia.org/wiki/Driveline

http://en.wikipedia.org/w/index.php?title=Power_controller&action=edit&redlink=1

http://en.wikipedia.org/wiki/Battery_%28electricity%29

http://en.wikipedia.org/wiki/Electric_motor



http://en.wikipedia.org/wiki/Electronic_throttle


6. The charging circuit through which the

regenerated energy is fed to the battery is

placed between the motor and the batteries.

7. A contact breaker is used to change modes

namely driving and charging modes as shown

in figure 5.

MOTOR SELECTION - The Motor was selected

by studying the various characteristics curves of

motor and also by calculating the

1. Maximum RPM that motor should have.

2. Voltage that can be generated for the

Maximum RPM.

If,

V max is max velocity of the bike

N max is max speed of prime mover (motor)

g is Output driven to input driver ratio

r is effective wheel radius

r

g

V

N65.2

max

max [Max speed of e-bike = 45kmph]

Maximum RPM of the motor, N max = r

gV max65.2

Voltage generated,

Ф = Flux / pole in Weber.

Z = Total number of armature conductors

P = Number of poles

A = number of parallel paths (= 2)

N = Speed of armature in rpm (N max of motor)

LAYOUT FOR OPTIMIZED CHARGING

Fig.6. Layout for optimized charging

Fig.7. Driving circuit for RBS

Fig.8. Charging circuit for RBS

Fig.9. Electronic Control Unit for charging circuit

MATHEMATICAL MODEL FOR

REGENERATIVE BRAKING SYSTEM

EFFICIENCY OF THE REGENERATIVE

BRAKING SYSTEM

Drag force ,

C d = Vehicle aerodynamic drag coefficient

ρ = Air density, Kg/m3

A = Vehicle frontal area, m2

V = Vehicle velocity, m/sec

Rolling resistance,

M = Vehicle test mass, Kg

g = Acceleration due to gravity, m/sec2

Cr = Rolling resistance coefficient

10


Acceleration force,

M = Vehicle test mass, Kg

a = Vehicle acceleration, m/sec2

Total power needed to propel the vehicle,

P = (F d + Fr + F a) V

Total energy required, E = ∫ P dt

Case (i) when a ≥ 0 (vehicle is accelerating)

E w = ∫ Pt dt

E w = Energy output at the driving wheels

Pt = Power at instantaneous time t

E1 = E w / η1

E1 = Energy input to the power train

η1 = Overall efficiency of the power train system

Case (ii) when a ≤ 0 & P ≥ 0 (vehicle slows down

or braking)

Battery is still supplying energy to maintain the

vehicle velocity

E w = ∫ Pt dt

E w = Energy output at the driving wheels

Pt = Power at instantaneous time t

E1 = E w / η1

E1 = Energy input to the power train

η1 = Overall efficiency of the power train system

When a ≤ 0 & P ≤ 0, the regenerative braking is

activated

Vehicle kinetic energy is converted to electrical

energy which is returning to the battery

Er = - ∫ Pt dt

E2 = Er / η2

Where,

E r = Energy regenerated

E2 = Energy returned to the electric storage system

η2 = Overall efficiency of the regenerative system

E2 / E1 express the percent of regenerative energy

returning to the vehicle electric storage system.

E1 expresses the energy consumption of the vehicle

without regenerative braking system.

CONVERSION OF KINETIC ENERGY INTO

USEFUL WORK

Stopping distance: the Distance travelled by the

vehicle from the moment the brakes are applied,

Braking Force: the Force required to apply the

brake or to stop the vehicle

U = initial velocity

V = final velocity

W = Weight of the vehicle

F = retardation

t = braking time or stopping time

g = 9.8 m/s2

By Newton equations of motion,

V = u - ft [Once brake is applied V=0]

f = t

u

fsUV 222

fsU 22

Stopping distance, S = f

u

2

2

Braking force, F = fg

w

(When the vehicle moves on a level road)

Work done = Braking force x Distance

moved

Work done = Heat generated

11


Heat generated = fg

w x f

u

2

2

= g

w x

2

2u

= 2

2

1u

g

w

= 2

2

1mu

The power absorbed by brakes during a stop can be

given as,

Braking power =t

EK

1000

.

(K.E Kinetic energy in J, t time, sec)

In general,

Power = timetaken

workdone

SIMULATION IN MATLAB

Variation Of Voltage Generated With Vehicle

Speed

Fig.10.Simulation of Voltage generated

From the result, it was found that the generation of

voltage is high at high speeds. So, the efficiency of

regenerative braking is more at high speed than at

the low speed. To improve the efficiency of the

bike at low speed, an ECU is developed for

optimized charging.

Variation Of Discharge Time Of The Battery

With Vehicle Speed

Fig.11. Simulation of Discharge time of the battery

From the above graph, it is inferred that the

discharge time of the battery decreases as the speed

of the bike increases. It is because the flow of

current to the motor is more at high speed than at

the lower speed.

EXPERIMENTAL VALIDATION

Fig.12. Testing the E-bike on chassis dynamometer

Work done = Kinetic Energy

Heat Generated = Kinetic Energy

12


The electric bike was tested for its battery

discharge rate. The vehicle was run on the chassis

dynamometer to measure the battery discharge rate

by subjecting it to numerous frictional brakes at

various speeds. Then the regenerative braking

system is used and similar tests are conducted to

find the improved performance in the battery

discharge time.

Also using the chassis dynamometer, the voltage

generated at various speeds and the stopping

distances are measured. Finally the efficiency of

regenerative braking technology is calculated from

the recorded values and the following characteristic

curves are drawn.

Vehicle speed vs. Voltage generated

Vehicle speed vs. Regeneration period

Vehicle speed vs. Deceleration

Vehicle speed vs. Battery discharge rate

VOLTAGE GENERATED FOR VARIOUS

VEHICLE SPEED

The D.C motor was tested without any load to find

its generated voltage at various speeds and the

readings are tabulated.

Fig.13. Voltage generated for various vehicle speed

Then the motor was fitted in the bike for applying

regenerative braking system in it and tested in the

chassis dynamometer and the graph is plotted for

generated voltage at various speeds of the bike and

it is shown in figure 13.

From the result, it is inferred that the amount of

voltage generated at low speed is less when

compared to that of higher speed of the vehicle. As

the speed of the vehicle increases, the generated

voltage also increases and it is very high at the

maximum speed of the bike.

STOPPING TIME (REGENERATION

PERIOD) AND DECELERATION OF THE

VEHICLE FOR VARIOUS VEHICLE SPEED

The bike is driven at various speeds on the chassis

dynamometer and their respective stopping time or

regeneration period are measured by applying the

regenerative braking i.e. by cutting down the

supply to the motor.

Fig.14. Variation of stopping time (regeneration

period) with vehicle speed

Fig.15. Deceleration of the vehicle with vehicle

speed

Vehic le s peed vs Voltag e g enerated

0

5

10

15

20

25

0 10 20 30 40 50

Vehic le speed (K mph)

Vo

lta

ge

ge

ne

rate

d (

V)

V normal

V load

Vehic le s peed vs S topping time

0

2

4

6

8

10

12

14

16

18

0 10 20 30 40 50


Sto

pp

ing

tim

e (

se

c)

S topping time

Vehic le s peed vs Dec c eleration

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0 10 20 30 40 50


De

ce

lera

tio

n (

m/s

ec

2)

Decc eleration

13


The deceleration of the vehicle was calculated for

each stopping distance and the graphs are drawn in

figures 14 & 15.

The stopping time or regeneration period increases

with the increase in the vehicle speed. The bike

takes more time to stop at higher speeds when

compared to lower speeds. This is due the velocity

of the vehicle which will be higher at high speeds.

BATTERY DISCHARGE RATE FOR

VARIOUS SPEEDS WITH AND WITHOUT

RBS

The vehicle was driven for various speeds on the

chassis dynamometer. First the bike was driven

without RBS by applying numerous brakes during

operating period. Then the bike was driven with

RBS by subjecting it to the same number of

braking (Regenerative braking). The battery

discharge rate was measured for both the cases.

The graphs are shown in figure 16.

Fig.16. Comparison of Battery discharge rate for

various speed with and without RBS

Battery discharge rate is very high at high speeds of

the vehicle and as the speed of the vehicle

decreases the discharge rate also decreases. So the

bike can run for more time at lower speeds than at

high speeds. With the application of regenerative

braking technology, there is a significant

improvement in the battery discharge rate.

CALCULATION OF EFFICIENCY

The improved average efficiency for the

discharging time of the battery in both cases (with

and without RBS) with respect to speed is

calculated as follows.

Where

T with RBS = Time for discharge rate with RBS

T without RBS = Time for discharge rate without RBS

Average calculated efficiency of the electric bike

with regenerative braking system with respect to

the speed was 9.05%

CONCLUSION

From the results, it was found out that the voltage

generated at high speeds is more and the battery

discharge rate is improved by Regenerative

braking. The efficiency of the battery was

increased by 9.05% and it was further increased

twice in sloppy areas. Also, an ECU has been

developed to utilize the voltage generated through

RBS to charge the battery. Moreover when

applying regenerative concept, the accompanying

friction (electrical resistance) assists the normal

brake pads in overcoming inertia and helps slow

the vehicle.

Future development can be done by replacing D.C

Motor by a SRM (Switch Reluctance Motor) for

improved overall efficiency and electronic speed

control can be introduced to control the speed of

the motor. Regenerating braking system can also be

applied in electric trains for improving its

efficiency.

ADVANTAGES

1. Acts as secondary braking system thereby

reducing the braking effort and increases safety.

2. The battery discharge is very less or sometimes

nil depending on the efficiency of regeneration.

3. The regenerative braking helps the drivers

enjoy ‗something for nothing‘.

4. For electric vehicles, the time for recharging the

batteries is considerably reduced.

Vehic le s peed vs battery dis c harg e rate

0

5

10

15

20

25

30

35

0 10 20 30 40 50

Vehic le speed (kmph)

Ba

tte

ry

dis

ch

arg

e r

ate

(m

in)

without R B S

with R B S

14


5. Increases the life of the Brake Pads.

6. Size of the batteries could be reduced if

regenerative technique is used for the vehicle.

7. Highly effective in hilly areas and slopes.

LIMITATIONS

1. The main issue with regenerative braking is

that it still relies on friction braking too.

2. Consequently the friction brake is still

necessary to bring the vehicle to a complete

halt.

3. Efficiency of Regenerative braking at lower

speed is less.

REFERENCES

1. Brent, Mark R. Papadopoulos & Jim M -

―Regenerative braking‖- United States Patent

4744577.

2. Jeffrey M. Christain - ―World guide to

battery powered road transportation-

comparative technical and performance

specifications‖, publisher George P. Lutjen,

McGraw- Hill publications.

3. Nakazawa. N, Kono. Y, Takao. E, Takeda.

N. - ―Development of a braking energy

regeneration system for city buses‖-SAE

paper 872265.

4. Satish C. Reddy and G. V. N. Rayudu -

―Design of regenerative braking system for

buses‖, University of Toledo, IIT Madras

paper 892529.

CONTACT

1 Student M.E. [Automobile Engg.],

2 Asst. Prof.,

Madras Institute of Technology,

Chennai

E-mail: [email protected]

ABBREVIATIONS

EVs Electric vehicles

RBS Regenerative Braking

System

D.C Direct Current

h.p Horse Power

rpm Revolution per Minute

V Voltage

i Current

T Temperature

A, amps Ampere

AH Ampere Hour

ECU Electronic Control Unit

15

National Conference on Advances in Automotive Technology [NCAAT 2010]

Rajalakshmi Engineering College, Thandalam, Chennai: 15th & 16th July 2010

N-2010-D03-03

Frontal Impact Analysis on a Heavy Passenger Vehicle

Jameer. M*, Dr. G. Devaradjane

Madras Institute of Technology

Keywords: Bus body, Front impact, safety, FEA,

ABSTRACT

Among all the accidents that take place, frontal

impact has got a major share of 40%. Again, in

these conditions the injury caused to the driver or

the front passenger is extremely high. In

automotive domain, more emphasis has been given

to the safety of passenger cars, but seldom is the

importance given to the passenger bus. Though the

damage due to frontal impact of the bus is lesser

when compared to other vehicles, the consequences

of such impact on drivers are fatal. According to

the study during frontal impact of bus more than

80% of drivers die than any other members of the

bus. In frontal impact scenario more significance

should be given to the structural integrity, and

hence this work is carried out in this direction. The

final solution of this project will be designing a

safe structure or a good structural integrity that can

withstand a frontal impact in which most of the

passenger and driver safety is assured.

INTRODUCTION

In today‘s growing population, there a need of

transportation to every passenger by any means.

Transportation means like two wheelers, cars,

buses, train etc. but cars and two wheelers are

mainly used by middle and higher order of the

society and also it has a very little passenger

capacity. The range that they are being used is also

limited to a small extent. For instance, no

passenger will try to commute a distance of 500km

or more in a car. This first raises the question of

safety and mainly depends on the driver‘s skill. In

this scenario, any passenger will opt for a

transportation medium which is safe, reliable and

cost effective. Upon regarding all the above

conditions, the highly used mass transportation

medium is the Heavy Passenger Vehicle, Bus. The

Bus has a good reliability compared to any other

vehicle and it also offers supreme comfort like

push back seats, neck support etc., in long journey.

Mainly it is cost effective when compared to that of

a car for the same distance or long journey. Most of

the journeys commuted in passenger bus ends in

safe manner. Despite of everything, some chain of

events cause failure of parts or some incidents that

may be fatal. There are many cases of causalities.

An impact with a static or dynamic object or a fall

from a datum height, etc. is some of the factors

which end in causalities. One of the main cases is

the frontal impact which is more dangerous than

any other situation. This definitely kills the driver

and the passengers in the front of the vehicle. If the

opposite vehicle has more kinetic energy (i.e. it

approaches in higher speeds), then the case will be

more fatal.

In automobile domain, safety features are mainly

considered for cars. The safety system in a car is so

worth full that, the passenger has a great chance of

evading death even the scenario of accident is on

the higher side. But this is not the same for a

passenger bus. This project has been directed in

that direction, which promotes safety to the driver

and the passengers which are seated at the front,

mainly in a frontal impact scenario. The scope of

the project is to build a structure which will tend to

protect the drivers and the passenger in a frontal

impact scenario.

WHY FRONTAL IMPACT

In this project we have considered the frontal

impact scenario as the main theme or scope. This is

because; the causality level to the driver and

passenger is more in frontal impact. Also there is

less case of severity when compared to side impact

or collision with a static member. Also in case of a

16



frontal impact, if the driver survives, then he can

help other injured passengers or passengers in need

of emergency care. So in this project frontal impact

has been taken as a main aspect.

NEED FOR BUS BODY DEVELOPMENT

Because of globalization, international players

have entered the bus market. It has created a

fierce competition.

Customer is more demanding and looking for

safety and comfort. It calls for high end

engineering capability for bus design.

Optimize design for contradictory parameters

viz, safety, reliability, weight,

manufacturability and cost.

Built in quality-in manufacturing calls for

detailed engineering efforts, modern

manufacturing facilities and good quality

control system.

Body builders are expected to become a

complete engineering industry with good

design capability besides manufacturing and

quality.

CRASHWORTHINESS

From an engineering perspective, crashworthiness

is the ability of the vehicle to prevent occupant

injuries in the event of an accident. This topic is

then the technical foundation for the legal doctrine

of crashworthiness or enhanced injury theory. It is

worth noting that the cause of the accident is

technically irrelevant in crashworthiness cases even

if the severity of the accident is an issue. Severity

can be assessed independently of the cause of the

accident. Severity is recorded in steel- in sheet

metal damage for the most part. Preceding

instances of human agency and even mechanical

failures that produced the record are not relevant

for interpreting it. If accident causation is an issue

in a crashworthiness case, it is for legal reasons

then and not technical ones.

Crashworthiness is not the same as vehicle safety,

and the two topics must be distinguished. The

safety afforded by a vehicle depends both on

crashworthiness and accident avoidance features,

the latter including such things as ABS, good

handling characteristics, or even oversize tyres.

These two concepts are frequently confused to the

detriment of those raising the crashworthiness

issue. One vehicle might be safer statistically than

another and still have a significant crashworthiness

defect. It could even conceivably be less

crashworthy overall while still being a "safer"

vehicle. This is because vehicle crashworthiness

depends on designed in features as well as

equipment specifications which can be viewed as

design features. A given vehicle either has these

features or it doesn't, regardless of its accident or

even injury rates.

FACTORS INFLUENCING BUS BODY

DESIGN

Exterior/Aerodynamics.

Weight and cost.

NVH and HVAC factors.

Layout and ergonomics.

Riding and handling.

Structure and Reliability.

Safety.

METHODOLOGY

Literature review on crashworthiness of bus is

carried out by referring reviewed books,

journals and related documents.

Geometric modelling of the bus structure is

carried out by using design software like

CATIA V5 R18.

FE model generation for all parts is carried out

using Hypermesh.

Input deck for simulation is created using

Hypermesh.

Frontal impact simulation is carried out using

LS-DYNA and post processing will be carried

out using LS-DYNA POST.

Investigation of the analysis result in order to

improve crashworthiness.

DEVELOPMENT OF BUS BODY KITS

There are six body kits which were developed for

the project. These are real time models which are

being built at bus body coach builders at Karur. A

blue print of the model has been acquired form

them and this model has been developed, as given

in figures from 1 to 6, using CATIA software. Kits

are

17



Floor kit.

Roof kit.

Left side kit.

Right side kit.

Front kit

Rear kit.

Figure 1. Floor kit

Figure 2. Roof kit

Figure 3. Left side kit

Figure 4. Right side kit

Figure 5. Front kit

Figure 6. Rear kit

Chassis of Ashok Leyland 210T is used. It is an all

steel structure. Afer the developmant of these

individual kits, the chassis kit is developed. Finally

all these kits are assembled to form a total bus body

structure.

18



ASSUMPTIONS MADE WHILE

DESIGNING.

Parts which are not directly related to frontal

impact or which has no significant effect on

the final output are not considered.

All the sub-systems which are discarded in the

design process has been considered as a

lumped mass at appropriate locations.

All structural design is as per the documents

obtained from PEETEE COACH BUILDERS,

KARUR.

Figure7. Assembled view

Figure8. Sectional view

DEVELOPMENT OF MATHEMATICAL

MODEL

The model which has been developed has to be

validated. But there is no sophisticated facility in

India to do a detailed work on this project. So there

are some other ways to validate the model. An

alternate way to achieve the proposed result is to

develop a mathematical model.

Macmillan (1983) proposed an alternative approach

based upon the results of many impact tests into

barriers. The acceleration, velocity, displacement

(crush) results of barrier impact tests tends to

display similar characteristics. The acceleration

curve has high frequency modulation caused by

erratic crumpling of the vehicle structure. The

velocity and displacement curves are progressively

smoother because of the filtering effect inherent in

integration. However these curves need to be

idealized in order to examine the overall behaviour

during impact and hence, in turn the effect of this

behaviour on the vehicle occupants. Macmillan

stated that what is needed is an analytical

expression for smoothened curves that satisfies the

following criteria,

It must be simple enough to be manipulated.

It must satisfy the boundary conditions found

in the curves in the impact tests.

It must correlate with the well-known test

cases and hence justify its use to predict the

outcome over a range of unknown examples.

It must be capable of representing the

behaviour of vehicles with different crush

characteristics with change to a small number

of variables.

The expression must also be applicable for all

values of e from 0 to 1 and must satisfy the

condition that

0dt

da

At, t = t2

This ensures that an instantaneous rate of change of

acceleration does not occur at the end of the

impact.

Macmillan proposed the dimensionless equation for

acceleration as follows:

Acceleration

2

1

22

1

t

t

t

t

t

cva

19



Where c is a dimensionless constant, to be

determined, and b is a dimensionless index greater

than unity.

Let 2

1

t

tT and integrate which becomes

v = v1 av(T)

where

c

eTT

c

Tv

a

2

2

11

1

1)(

Substituting for av (T) and integrating

Assumption :

Collision is plastic

e=0 Large displacement (v2=0)

V= v1av(t)

At t = t1=0, v=v1

So av(t) = 1

ß0 = 2

The parameter ß0 is called the structure index

because soft nosed vehicles have small values of ß0

and hard-nosed vehicles have larger values. A

typical value for a medium sized car is ß0 = 2.

When the mean force is small, the impact is almost

elastic (e-->1), and when it is large the deformation

is almost plastic (e0).

c

eTT

c

Tv

a

2

2

11

1

1)(

At t= t1=0

02

0)(

tTLet

02

1

1

11

c

4

1

3

1)(

c

Tv

a

12

1)(

c

Tv

a

c= 12

)2

1( xtt

Let

2*

2

2

1

11tdx

XXCVS

dxtdt

2

2*tdxdt

dtc

ecvvdt

2

2

2t

t1

2t

1t

1

1

2t

t1

1

2t

1t

2*

2

2

1

11tdx

XXCVS

0

2

1

3

32

2

2121

t

t

XXCtVS

32

1

21

1CA

A

CTS

a

32

1

21

1

2

32

2t

t1

1

21

2t

t1

)(

Displacement S = v1 t2 as (T)

The velocity, displacement and acceleration are

calculated for the speed of 25, 30 & 35 Km/hr by

using the above equations and the results are

tabulated and plotted on the graph.

Assumption e=0 (v2=0)

20



ANALYTICAL RESULT - 25 KM/HR SPEED

OF IMPACT

The velocity, displacement and acceleration are

calculated at the speed of 25Km/hr and the results

are tabulated in Table shown below.

Time

(s)

Acceleration

(g)

Velocity

(km/hr)

Displacement

(m)

0 0.00 25.000 0.000

0.02 4.15 23.029 0.135

0.04 6.10 18.457 0.251

0.06 6.36 12.970 0.338

0.08 5.42 7.813 0.396

0.1 3.81 3.790 0.427

0.12 2.03 1.270 0.441

0.14 0.59 0.177 0.444

0.16 0.00 0.000 0.444

Graph 1. Time vs Accleration

Graph 2. Time vs Accleration vs Displacement

DEVELOPMENT OF FEA MODEL

The conventional model which was developed in

CATIA software has to be meshed for analysis of

crash. For this HYPERMESH software is used. Altair HyperMesh is a high-performance finite

element pre-processor that provides a highly

interactive and visual environment to analyze

product design performance. With the broadest set

of direct interfaces to commercial CAD and CAE

systems, HyperMesh provides a proven, consistent

analysis platform.

Figure9. Meshed Bus body model

Steps involved in Meshing.

Geometric cleanup.

Taking mid surface.

Rough mesh and Quality check.

Applying contact elements.

Rigid surface for crashing.

Creating control card for crash.

Exporting the FEA model to LS Dyna

software.

PROCESS INVOLVED IN EXPORTING THE

MESHED MODEL TO LS DYNA.

• Checking connectivity.

• Spot welds and rigid connections has to be

specified

• Checking connectivity between elements.

• Combining nodes of every structure.

• Impact area definition.

• Development of rigid area.

• Material selection for rigid element.

• Creation of rigid obstruction.

• Creating contact card.

• Surface to surface contact has to be

defined.

• Defining master and slave contact cards.

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18

TIME (s)

AC

CE

LE

RA

TIO

N (

g)

Acceleration(g)

0

2.5

5

7.5

10

12.5

15

17.5

20

22.5

25

27.5

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16

TIME (s)

VE

LO

CIT

Y (

Km

/hr)

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

DIS

PL

AC

EM

EN

T (

m)

VELOCITY (Km/hr)

DISPLACEMENT (m)

21



• Defining AUTMOVE option.

• Velocity of contact cards has to be

defined.

• Velocity of particular nodes has to be

mentioned.

• Control Energy Development.

• To obey cube rebound theory.

• Shell element and solid element internal

energy definition.

• Defining Control output.

• Result frequency has to be defined.

• Control termination.

• Defining end time of the impact or

process.

• Number of cycles to be defined.

Figure10. LS Dyna interface.

The meshed model is imported to LS Dyna

software for crash analysis. The conditions are

Velocity of the vehicle: 30km/hr.

Crash type : frontal impact

Obstruction: rigid barrier.

Simulation time: 0.02 sec.

Figure11. Output window

RESULT FOR CONVENTIONAL MODEL

CRASH AND DEVELOPMENT

The conventional crash test run showed that there

exists a maximum energy level of 1.6986x107

KN/cm2. This much of energy is involved in the

frontal impact scenario. This kind of energy level

may bring harm to the front seated passengers and

driver too. So this energy must be damped before

reaching the frontal section. As a part of design

change, crash initiators have been developed.

These are structures having the same material

which possess some design change to improve the

crashworthiness. There exists more number of

structures, but some of the chosen one is shown

below, which was developed in CATIA software.

Figure 12: Crash Initiators

An ANSYS analysis has been done on these four

types of crash initiators. The conditions are

All crash initiators having same length

according to the conventional beam.

All initiators will undergo a common load of

300KN/cm2

for one second.

The crash initiator which absorbs more energy

level is considered for design change.

When the analysis was done, V shape notch type

crash initiator absorbed more energy levels. So this

initiator was used for design change. When

replaced with the conventional model and a crash

analysis was done, the results are being positive.

Nearly 1.12% of the total energy was damped

when compared to conventional design.

1.12% here is nearly equals to 0.0191x107

KN/cm2.

22



Figure 13: comparison of energy levels

CONCLUSION

• It has been understood that the load

distribution on the structures is not uniform,

which lays down the road to improvement in

buckling characteristics of the structures.

• By having crush initiators, the peak load can

be reduced. This has been achieved by

implementing such designs to some of the

structural members, which is around 4%

reduction in peak load.

• The design improvement that has been

achieved is just for a few structural elements.

If this approach is followed for many other key

structural members then the design could be

far superior.

FUTURE WORKS

In today‘s automotive body engineering

advancements there are various systems that

improve the crashworthiness of the vehicle

significantly. Many such systems can be

implemented in order improve the structural

safety.

Simulation of the frontal impact behavior of

the passenger bus can be carried out by

considering various subsystems of the vehicle

like engines, transmission, steering system etc.

Positioning of dummy in the driver‘s seat

helps in finding the injury parameters.

Seat belt concept in passenger bus is an alien

concept in India; efforts can be made in

developing such a concept.

More understanding is required in order to

improve the structural behavior of chassis,

which can be detrimental in overall design.

ACKNOWLEDGEMENT

The authors gratefully acknowledge the support

given to this research by head of the department of

automobile engineering (Madras Institute of

Technology, Chennai-44).

REFERENCES

1. Manjunath Rao T.S., ‗Study of frontal impact

of a passenger buses‘ Coventry University.

2. ‗Driver and Crew Protection in Frontal Impact

on Bus‘. Science and Vehicle Conference,

Hungary Belgrade, 15-16 April, 2009

3. Tomas W. Tech; Ignacio Iturrioz & Inacio B.

Morsch. ‗Study Of A Frontal Bus Impact

Against A Rigid Wall‘.WIT Transactions on

Engineering Sciences, Vol 49

4. Impact Loading of Lightweight Structures, M.

Alves & N. Jones (Editors) 2005 WIT Press,

5. A.E. af Wåhlberg . ‗The Stability of Driver

Acceleration Behavior, And A Replication Of

Its Relation To Bus Accidents‘.

www.sciencedirect.com/occupantsafety

6. Alejandro Palacio, Giuseppe Tamburro,

Desmond O‘Neill, Ciaran K. Simms ‗Non-

Collision Injuries In Urban Buses—Strategies

For Prevention‘ www.scridb.com

7. Simulation Of Crash Tests For High

Containment Levels Of Road Safety Barriers.

M. Borovinsˇek , M. Vesenjak, M. Ulbin, Z.

Ren. www.sciencedirect.com

8. Vincze-Pap Sándor Autókut, Csiszár András

Edag ‗Real And Simulated Crashworthiness

Tests On Buses‘. Hungary Ltd.,

www.nhtss.com/feamodeling/impacttest

CONTACT

* Jameer. M, II M.E Automobile engineering,

Madras Institute of Technology,

Chennai – 600 044

Email id: [email protected]

23



N-2010-D04-04

Finite Element Analysis of Inlet Manifold of an IC Engine

1 Karthikayan. S,

2 Sankaranarayanan. G,

3Ranganathan.L

Keywords: FEA, Inlet manifold, IC Engine, Swirl, Turbulence

ABSTRACT

Diesel powered vehicles are the most preferred

transport due to its performance and load carrying

capacity. The diesel engine combustion should be

efficient only then the power out would be

maximum. The better fuel efficient, high power

output and lower fuel consumption can be obtained

by various methods the combustion efficiency is

mainly decided by the gas motion within cylinder.

It is worth noting that the induction system and

mixing of fuel and air plays a major role and

efficient mixing of fuel with air is achieved by

creating a swirl in the chamber. The swirl may be

generated by redesigning the piston, valves or ports

and manifolds. This study was focused on

improving the mixing of air and fuel by suitably

modified manifold. The study has been done in

stages such as modelling, analysing with different

geometric for optimization of the manifold and

verifying the result experimentally. This paper

explains the study of the modelling and analysis of

manifolds with different helix angle using Pro/E

and CFD analysis using Element ―Flotran 142‖ in

CAE Software, ANSYS.

INTRODUCTION

The engine intake process governs many important

aspects of the flow within the cylinder. In four

stroke engines, it is the characters of the turbulence

at the end of the compression process that is more

important and it controls the fuel air mixing and

burning rates.

The flow through intake valve or port / manifold is

responsible for in-cylinder flow characteristics.

When a swirling flow is generated during intake, an

almost solid body rotating flow develops which

remains stable for much longer than the inlet jet

generate rotating flows.

Swirl is defined as organized rotation of charge

about the cylinder axis2. Swirl is by bringing intake

flow into the cylinder with an initial angular

momentum. Swirl generation is done during

induction. In a typical swirl motion, the flow is

discharged into the cylinder tangentially towards

cylinder wall where it is deflected sideways and

downward in a swirling motion. In the other swirl

is generated largely within the inlet port, the flow is

forced to rotate about the valve axis before it enters

the cylinder.

Nico Ladommatos et.al have analysed and

estimated the effect of the swirl in unsteady flow

during intake process by modifying the swirl with

Bowl in piston combustion chamber. The study has

involved finding the effects by varying the injector

holes and simulation was done to predict the swirl

and cross wind velocity.

Jun-ichi-Kawashima et.al3 have conducted a

modelling analysis and verified experimentally

using small high speed 4 valve DI diesel engine.

Swirl ratio was measured with an impulse swirl

flow meter (Vane method).

Lawrence W. Evers has studied the characteristics

of the transient spray from a high pressure swirl

injector. The swirl injector sprayed small droplets

with large cone angle and evaporated quickly and

reduce wall wetting which is desirable for stratified

charge engine. It has been found that increasing the

swirl increased the droplet size in leading edge and

cone regions. And increasing the fluid pressure

reduced the size of the droplets in the leading edge.

24



MODELLING

Initially the straight inlet port was modelled using

Pro/E. The port was having different cross section

along its path. The Path was taken as a Trajectory.

It is a 3D curve drawn by two projection method

using curve option. In the two projection method

the front view of the 3D curve can be drawn on the

front plane and top view of the 3D curve can be

drawn on the Top plane finally getting the 3D

curve. Along the 3D curve the points were created

wherever we need the different cross section using

offset coordinate system in datum points option

from the datum features. A sweep is created by two

sections. The first session is the trajectory and the

second is the cross section. Trajectory is the path

along which the cross section is swept but the cross

section should be uniform.

MESH GENERATION

In order to analyze the effects of the intake

manifold geometry and examine measures for

making further improvements, it is essential to

know the flow states in the manifold and in the

cylinder. Therefore, three - dimensional flow

simulations were conducted to investigate flow in

the intake manifold. Because many different

variations are conceivable for the helical manifold

geometry, an automatic mesh generator was

developed that can produce computational meshes

with high efficiency. Approximately 30 parameters

are fed to the generator to define the manifold

geometry and the computational mesh divisions are

then generated automatically. Cylindrical

coordinates are used to generate a homogeneous

mesh in the circumferential direction. Emphasis is

given on the quality of the mesh around the valve

and the valve seat, which greatly affects flow

characteristics.

The cross – sectional shape of each part of the

helical manifold is defined on the basis of

parametric calculations, and the mesh topology is

then generated by arranging the shapes in the

cylindrical coordinate system.

PROBLEM DEFINITION

In this paper the turbulent mixing of fuel and air in

the engine was improved by redesigning the inlet

manifold. Actually the engine has a two manifold

inlet and a straight outlet manifold located at the

cylinder head. The air coming through the straight

inlet manifold to cylinder was of laminar type. The

vector velocity of the inlet air particles was

vertically downward in direction.

During the injection, the fuel is sprayed vertically

downwards. Due to downward movement of the

fuel and air, there is low relative velocity between

them inside the combustion chamber. In order to

get better mixing, the turbulence or swirl is to be

created inside the combustion chamber. Hence the

inlet manifold was modified into helical manifold.

APPROACH

The flow through the inlet manifold was performed

using an analysis tool ANSYS CFD using Flotran

142 element.

Similarly the modified inlet manifold or helical

manifold was modelled using Pro/E and the flow

will be analyzed by using ANSYS CFD with same

element Flotran 142.

Figure1. Wire frame model in straight manifold

Figure2. Swirl manifold model in 160°

25





Figure5. Meshed model in Swirl Manifold 160°

RESULTS AND DISCUSSION

Swirl manifold:

In the case of 160o swirl angle, the swirl factor was

found to be 0.26. It was seen that many streamlines

of the incoming air followed the cylinder wall and

generated strong swirl. With a value of swirl angle

of 2200 the swirl factor was found to be 0.535. On

the other hand, a larger proportion of the

streamlines were directed toward the centre of the

cylinder. With a value of swirl angle of 300o the

swirl factor was found to be 0.424.

Figure6. Vector plot in swirl manifold 160°

Figure7. Contour plot in swirl manifold 160°


These results can be understood in terms of the

flow patterns shown in Fig 11. The rotation in the

figure is a schematic representation of the flow

patterns for different values of swirl angle. Intake

air flowing through the straight portion of the

26



manifold strikes the scroll, which changes the flow

direction. The air then swirls as it flows along the

wall towards the valve seat. Consequently, the

direction of the outflow from the valve seat is

determined by the swirl angle imparted to the

intake air flow in the throat of the manifold.

Increasing the height from the tube to the valve seat

imparts a larger swirl angle to the intake air in the

throat, causing a larger portion of the incoming air

to flow towards the centre of the cylinder.

Table 1 - result in swirl port 160°

Swirl factor = (√VX2+VY

2/VZ)

= (√5.77E+062+97505192 /2.67*106)

= 0.424


Figure9. Vector plot in swirl manifold 220°

27




Figure 11 Vector plot in swirl manifold 300°

Straight manifold

The straight manifold has a small swirl factor, so

that the mixing will be very less. The manifold

mesh used in the calculations had approximately

70,000 elements, including a 100 mm long cylinder

and the plenum provided at the manifold inlet. A

second-order upwind finite differencing scheme

was used in the calculations. The scope of the

following analysis will be limited to just the

qualitative flow patterns. Intake air flowing through

the straight portion of the manifold then bends in

the bent portion of the manifold and hits the valves

and gets spread. But as the air moves straight into

the block the problem persists as of during the

compression stroke the air is compressed and the

piston travels opposite to the flow hence retarding

the momentum of the incoming air. This reduces

the mixing of air and fuel and hence leaving un-

burnt fuel in various regions.


A swirl factor of 1.568 is very low swirl factor

which likely to effect the performance of the

engine.

These aspects give us a way to introduce the swirl

in the scroll portion which can increase the swirl

that would improve the mixing of air and fuel in

the engine block.

28



Figure12. Meshed model in straight Manifold

Figure13. Vector plot in straight manifold

It could also be seen that the vertical velocities at

the manifold ends are also poor hence further

retarding the performance. There are also seen

small whirls which are likely to decrease the

performance.

Table 4 - result in straight port

CONCLUSION

From the analysis it is observed that the swirl

factor for 220° is higher than the one obtained

from the 160° or 300°. Hence 220° swirl angle

would be the optimum which will improve the

engine performance and is almost having 100%

better performance compared to the 160° degree

swirl angle.

The turbulence in the 300° model has been

identified to be on the higher side compared to

the 160° and 220° swirl angles. This could be a

reason for the decreased value in the swirl

factor.

The turbulence has been identified to be more in

the bending portion.

Intake air flowing through the straight portion

of the manifold strikes the scroll, which changes

the flow direction.

While change in swirl direction it was identified

that the 300° swirl angle mode has small

reversed eddies. The air then swirls as it flows

along the wall toward the valve seat. These

portions do not seem to affect the swirl of the

intake air.

In the 220 and the 300 models, increase in the

height from the scroll to the valve seat imparts a

larger swirl angle to the intake air in the throat,

causing a larger portion of the incoming air to

flow toward the centre of the cylinder.

Due to reasons mentioned above, the 300° swirl

angle reduces the angular momentum in the

cylinder. This results in a smaller swirl ratio.

This also plays a major role in reducing the

swirl factor in the 300° case compared to the

220° swirl angle.

DEFINITION

Swirl: Organized rotation of the charge about the

cylinder axis.

Swirl ratio: Ratio between the angular velocity of

the solid body rotating flow (ωs), which has equal

angular momentum to the actual flow, to the

crankshaft angular rotational speed. Swirl Ratio

(Rs) = ωs / 2πN.

29



ABREVIATIONS

3D – THREE DIMENSIONAL

CAD – COMPUTER AIDED DRAWING

CAE – COMPUTER AIDED ENGINEERING

CFD – COMPUTATIONAL FLUID

DYNAMICS

DI – DIRECT INJECTION

SAE – SOCIETY OF AUTOMOTIVE

ENGINEERS

REFERENCES

1. Chen & Dent J. C. (1994), ―An

Investigation of Steady Flow through a

Curved Inlet Port‖, SAE paper volume II

950818

2. John B. Heywood, ―Internal Combustion

Engine Fundamentals‖, McGraw Hill

International Editions

3. Jun-ichi kawashima, Hiroshi Ogawa and

yoshiyuki Tsuruhave (1998) Research on

a Variable Swirl Intake Port for 4-valve I-

speed d1 Diesel Engines. SAE Paper

982680

4. Justin Seabrook and Mike Fry, Cosworth

Technology, Ltd. ―Emissions and

Performance of a Carbon Fibre Reinforced

Carbon Piston‖ SAE paper 2000-01-1946

5. Lawrence W. Evers, Michigan

Technological Univ, ―Characterization of

the Transient Spray from a High Pressure

Swirl Injector‖ SAE paper 940188

6. Safdari. Y. B. and K. Kumarasekaran,

Bradley Univ ―A FEM Thermal Analysis

on a Novel Designed Air-Gap Insulated

Piston‖ SAE paper 932490

7. Shuji Kimura, Yukio Matsui and Masao

Koike, ―A New Combustion Concept for

small DI Diesel Engines – 2nd Report:

Effect of engine performance-―,

Transactions of JSAE, Vol.28, No.2

(1997), 29(in Japanese).

8. Dr. Stanley K. Widener (1995),

―Parametric Design of Helical Intake

Port‖, SAE Paper 950818

9. Tippelmann G ―A New Method of

Investigation of Swirl Ports, ―SAE paper

770404.

10. Yasushi Mase, Jun-ichi Kawashima,

Tatsuo Sato and Masakazu Eguchi

―Nissan‘s New Mulivalve DI Diesel

Engine Series,‖ SAE paper 981039

11. William Church and Farrell P.V. (1998),‖

Effect of Intake Port Geometry on Large

Scale in – Cylinder Flows‖, SAE paper

980484.

12. Yukio Matsui, Jun-ichi Kawashima and

Kunihiko Sugihara ―Analytical Study of

Interaction between Combustion Chamber

Specifications and Engine performance in

DI Diesel Engines‖, Transactions of

JSAE, No .40 (1989), 34(in Japanese)

13. Yukio Matsui, Shuji Kimura and Masao

Koike, ―A New Combustion Concept for

small DI Diesel Engines - 1St Report:

Introduction of the basic Technology-―,

Transactions of JSAE, Vol.28, No.1

(1997), 41(in Japanese)

CONTACT

1 Research scholar, Sathyabama University,

Chennai – 119 [email protected]

[email protected]

2 Professor, Adhi Parasakthi Engineering College,

Melmaruvathur – 603 319

[email protected]

3 Anna University, Chennai – 25

30



N-2010-D08-10

Roof Mounted Driver’s Seat for Armoured Vehicle

J. Jaganathan1, Dr. Velamurali

2, V. Balaguru

3

Keywords: Armoured vehicle, driver seat

ABSTRACT

This research work focuses on Design and

Development of Roof Mounted Driver‘s Seat for an

Armoured Vehicle of Main Battle Tank. The Driver

plays a significant role in tactical manoeuvring

through war field. It clearly indicates the

significance of the role of Armoured Fighting

Vehicle driver in a combat situation. Hence, it is

obvious that the Tank driver comfort & protection

is one of the prime factors in increasing the combat

effectiveness of any Armoured Fighting Vehicle.

The present MBT Driver is provided with 1.2m3

space, where all the driving controls &

instrumentations are placed for effective control of

the Fighting Vehicle .The driver‘s seat is mounted

in the driver‘s compartment directly below the

elliptical driver‘s hatch. The seat is secured on the

floor plate with fasteners. Also anti mine plates are

provided below the Driver‘s seat to protect the

driver from land mines.

INTRODUCTION

The purpose of the Roof Mounted Driver‘s Seat for

an Armoured Vehicle of Main Battle Tank is to

protect the driver from land mines. The roof

mounted driver‘s seat was designed considering

five aspects of ergonomics: safety, comfort, ease of

use, performance and aesthetics. The structure was

designed in order to withstand dynamic load and

any impact load during mine blast. It was fitted in

roof of the hull in tank, so that the crew will be in

safe and comfort working condition.

The Indian army needs to ensure survival of crews

from mine blasts which varies from 4 kg to 25 kg.

Hence, in depth study is required in the

development of driver‘s seat for better crew

comfort against terrain disturbances & enhanced

survivability against mine blast. Hence it is

proposed to develop a roof mounted driver seat,

where the driver suspends from the vehicle ceiling

like parachutists in their safety harness, thereby

This helps in attenuation of ground induced shocks

more effectively, when compared with floor

mounted seats.

Mine blast protective seat combines shock-

absorbing material to attenuate blast forces with a

steel plate to reduce shrapnel penetration combined

with a 2 point roof mounted brackets and adjustable

angle spine support. The suspension system also

combines a shock absorber / gas spring to reduce

gravitational forces. Mine blast protective seat uses

multi– density foam to reduce cramp and other

injuries during sustaining periods while in the seat.

DESIGN PHASE

Ergonomics is the scientific discipline concerned

with designing according to human needs and the

profession that applies theory, principles, data and

methods to design, in order to optimize human

well-being and overall system performance.

There are five aspects of ergonomics: safety,

comfort, ease of use, performance and aesthetics

1. Safety – E.g.: The provision of Seat belts,

which ensures the driver safety during

collision.

2. Comfort – E.g.: The viscoelastic cushion of

varying density provides the driving comfort

even after sustained cycle of operation

3. Ease of use – E.g.: The use of various

adjusting mechanism for back & forth

movement , also rising and lowering

mechanism and reclining mechanism

31

http://en.wikipedia.org/wiki/Scientific_discipline

http://en.wikipedia.org/wiki/Design



4. Performance – E.g.: The difference between

the actual vibration level at seating position

and the input at hull floor plate

5. Aesthetics – E.g.: The way the driver seat

looks such that the product speaks itself for

its quality.

When a conventional driver seat, which is floor

mounted, is subjected to mine blast, the generated

shock waves get transmitted directly to spinal

column and critical organs of the occupant, leading

to risk of fatal trauma.

Hence it is proposed to develop a roof mounted

driver seat, where the driver seat is suspended from

the vehicle ceiling like parachutists in their safety

harness, thereby attenuation of ground induced

shocks is more effective when compared with floor

mounted seats. Mine blast protective seat uses multi

– density foam to reduce cramp and other injuries

during sustaining periods while in the seat.

Necessities for the new concept of RMDS

An anti-tank mine, is a type of land mine designed

to damage or destroy vehicles including tanks and

other vehicles. Modern anti-tank mines are usually

more advanced than simple containers which are

filled with explosives detonated by remote controls

or the vehicles pressure

Mounting seats from the roof of the vehicle, rather

than the floor, will help protect occupants from

shocks transmitted through the structure of the

vehicle and a four-point seat harness will minimize

the chance of injury if the vehicle is flung onto its

side or its roof - a mine may throw a vehicle 5 - 10

m from the detonation point. The effect of mine

blast of 5kg TNT would cause death of the driver.

Even plate of 50mm thickness is deformed heavily.

Most modern mine bodies or casings are made of

plastic material to avoid easy detection. The mines

are buried in ground to a depth of 250mm. They

feature combinations of pressure and magnetically

activated detonators to ensure that they are only

triggered by vehicles.

Modelling of structure

Free body diagram of the various members are

drawn and analytical calculations have been made

for the reaction forces and moments. The structure

has been modelled in Pro-E software and numeric

calculation has been evaluated by vectorial method

and compared with the results with Analysis

software.

Fig.1 Layout of the Roof Mounted Driver‘s Seat

Actual Vertical Force is considered as 2000 N i.e.

Mass of the driver is 800N, Mass of the driver seat,

which is as shown in Fig1, is 500N and Mass of the

driver seat structure is 700N.

For safer side, the load is considered higher value,

the vertical load due to mass of driver, mass of

driver seat and mass of driver seat structure as 3g

(6000 N), horizontal load due to deceleration force

of vehicle as 2g (4000 N), sideways force due to

negotiating a turn as 1g (2000 N). In order to

convert the Dynamic analysis to equivalent static

analysis as per thumb rule 3g (6000 N), 2g (4000

N), 1g (2000 N) were taken in vertical, horizontal

and side wise respectively.

FATIGUE ANALYSIS OF STRUCTURE

The structure has been evaluated for fatigue

analysis for infinite number of cycles and found

within endurance limit.

Selection of fasteners

Bolts were designed according to loads acting on

them in LH and RH frames connected with bottom

plate. The shear loads is taken as 6000N and 1000

N acting at the each bolt. M12 standard bolts,

32

http://en.wikipedia.org/wiki/Land_mine

http://en.wikipedia.org/wiki/Tank



according to the IS: 1367 are used based on the

calculations.

Size of the bolt

We know that the equivalents tensile load (Wte) -

from that we can calculate the Core diameter of the

bolt.

Core diameter (dc) is calculated to be 10.118 mm

The standard core diameter is 10.118 mm and

preferred equivalent size of the bolt is M12 as per

the IS 1367-1979.

Shear pin

Shear pins are used to share the maximum shear

stress acting on the bolts to avoid failure. We have

selected the shear pin and based on the design

calculation, we preferred the 16mm shear pin

according to the IS 2393. The material of the pin is

C30 to IS 5517-1972 steels of hardening and

tempering or any other suitable steel with minimum

tensile strength of 500 N/mm2.

Fig.4.3 Shear pin for fixing the RMDS Bracket

Wte = (π/4) (dc) 2

σt

10x103 /4 = (π/4) (dc)

2 x100

dc = 5.64 mm

Sway pin

Sway pin is designed to withstand mainly the

bending stress. Allowable stress is is assumed to be

150N/mm2 (50 % of yield stress).

Diameter of sway pin is calculated as follows:

Bending Stress (σ b) = 150 N/mm2.

σ b = Bending Moment/Section Modulus

= M / Z

150 = 4472x62/ (П d 3/ 32)

Diameter of sway pin (d) = 26.6 mm

Sway pin is welded in the floor plate to arrest the

swaying of driver seat along the x and y directions,

for that we found the welding thickness by

following equation.

Max allowable weld leg size =12 mm (because

plate thickness =18mm)

σ = 5.66Mb / hD2 П (PSG DATA book)

45 = 5.66 * 4472.135 * 62 / (12x0.707) d2 П

For fillet (leg) size of the weld (h) = 8.484 mm

Diameter of pin = 36.66 mm

40mm diameter pin is preferred for welding

purpose.

Sway pin diameter = 40 mm.

Factor of safety of sway pin

We found the factor of safety for sway pin as

calculated by the following equation. EN 24 is

selected for sway pin.

Ult. Tensile Stress (UT) = 500 N/mm2.

Yield Stress (σy) = 300 N/mm2.

Endurance Stress (σe) = 250N/mm2.

Max.BM (M max) = Px62

= 277272.92 N mm

Min.BM (M min) = - 277272.42 N mm

Goodman method

1/FS = (0/500) + (45/250)

Factor of a Safety (FS) =11.11

33



Soderberg method

= (0/ 300) + (45 / 250)

1/FS = 0.18

Factor of a Safety (FS) = 5.55

CONCLUSION

Based on the confidence gained in wooden Mock-

up, design calculation and analysis results from

ANSYS workbench, we have fabricated a metallic

prototype and fitted on the MBT vehicle. The main

objective of the project is design and development

of Roof Mounted Driver‘s Seat is evaluated. Design

Concept for Roof mounted Driver‘ seat is suggested

for Main Battle Tank. Various Forces & Moments

acting in the roof mounted river‘s seat are

calculated and found to be less than the Von Mises

Stress. Design of the Structure for Roof Mounted

Driver seat is sturdy and safe.

REFERENCES

1. Ajeya, T72 Tank 1972 Technical descriptions

& operating Instruction Manual

2. Ajeya, T72 Tank 1972 [4] –Maintenance

operating Instruction Manual for crew

3. Bhisma, T-90S Tank 1990– Technical

descriptions & operating Instruction Manual

4. Bhisma T-90S Tank 1990–Maintenance

operating Instruction Manual for Crew

5. Janes Defence Journal (2009) for armoured

vehicle

6. www.global security.org

7. www.gowelding.org

8. Dr.Sadhu singh (1999) Strength of materials,

41-44

9. Norton (2002) Machine design.

CONTACT

1College of Engineering, Guindy, Chennai

[email protected]

2Head & Professor I/C, Department of

Mechanical Engineering, Anna University,

Chennai

3Scientist ―F‖, Head, Main Battle Tank,

CVRDE, Chennai

34



N-2010-D05-11

Parametric Study Analysis and Design Modifications of

Rear Axle Housing Assembly for Heavy Commercial

Vehicles

Ravichandran.G.1, Malarmohan.K.

2 and Chinnaraj.K

3

Keywords: Rear Axle Housing, design modification, stress analysis, Weight Reduction

ABSTRACT

Automakers are giving utmost importance to reduce

vehicle weight by engineering them for greater use

of lighter, stronger components. This project work

also aims at weight reduction of rear axle housing

assembly used in Heavy Commercial Vehicles

through design optimization or modifications. The

static and dynamic behaviours of existing design

rear axle housing are analyzed using ANSYS finite

element solver under major load cases like

acceleration, braking and cornering. After studying

the induced stress distributions in the existing

model, possible geometric locations for weight

reduction are identified. Subsequently, concept

models were generated considering the parameters

like wall thickness, shape and material for design

modifications. A quasi-static approach that

approximates the dynamic behaviour into static

equilibrium was followed to carry out the numerical

simulation. For 3D Solid Modelling, Meshing and

Structural analysis CAE software were used. After

analyzing the results of the proposed models, a new

model which can meet the existing stress levels

with considerable weight reduction is selected.

INTRODUCTION

The rear axle housings are stationary members

enclosing the rotating shafts i.e. axle shafts that

transmit power to the wheels. These semi-floating

axles are supported by bearings from inside the rear

axle housing. Rear Axle housing is used in the

Multi-axle Vehicles of Heavy Commercial Vehicles

such as Trucks, Tippers, and Transit mixture

carriers.

Both the rear axle assemblies of the multi-axle

vehicle are live axles. A small self-adjusting

propeller shaft receives power from the forward

rear axle and transmits to the rearward rear axle.

The Suspension comprises of a Leaf spring

assembly fitted on either side of the rear axle

housing. The mid-point of the leaf spring assembly

is attached to the chassis by a leaf spring holder

while the two ends are resting over the spring seats

on the forward and rearward rear axle housings.

There are totally six tie rods, three each for both the

rear axle housing. One tie rod is attached to the

chassis cross member and the top bracket of the

forward rear axle Housing. One tie rod on either

side (LHS & RHS) is attached to the leaf spring

holder and the bottom bracket of the forward rear

axle housing. Construction is similar for the

rearward rear axle housing also. The rear axle

housings have circular portion in the middle to

accommodate differential gear mechanisms. Either

side of this, the cross-section is rectangular with

good corner radius. Both the ends of the rear axle

housing are welded to the machined shafts. At both

the ends of the rear axle housing shafts are fitted

with bearings close to wheel track. On these

bearings wheel hub will be mounted along with

wheel and tyre. The brake assembly is also fitted on

the rear axle housing at suitable locations.

OBJECTIVE AND MODEL GENERATION

The assembly consists of 12 sub-parts as shown.

The objective of this study is exploring the

possibilities to reduce the weight of the assembly

by optimization or by generating alternate designs.

35



After analysis of the existing model, the induced

stress levels are found to be high. So, the generation

of alternate designs initiated. To do this design of

experiments is considered. By this way a proper

scientific approach to generate models and the

critical factors that contribute to weight reduction

can be obtained.

Cross-sectional shape, thickness and material are

the three parameters considered for two levels (low,

high). Thus, two level three factorial DOE is used

and so eight models were generated as given below

to achieve the objective. The shafts are critical

functional parts and they remain unchanged due to

fitment perspective. The dome, ring were retained

without modifications. However, spring seats,

bottom brackets and top brackets were modified

with new design. Also, the half-pressings were

modified for shape and thickness. Thus, the weight

of the rear axle assembly is reduced by analysing

part-by-part basis. Please refer the weight reduction

table 3 to find out the percentage of weight

reduction.

The Concept Development process followed is

similar to the sketch as given below in Fig 2.

Please note that further the experiment models will

be noted as ‗C‘ for circular, ‗S‘ for square followed

by the cross-sectional wall thickness. e.g. Model

‗C16‘ refers to Circular cross-section with 16 mm

wall thickness. Table 1 shows the critical

parameters under consideration used to develop the

concept models and the materials, thickness

considered. The thickness change was done in the

housing. Density was taken as 7850 kgm/m3

and

Poisson ratio as 0.3.

MESHING

Axle housing and sub-parts are so thick that we

can‘t use shell element. Therefore, 3D models of all

these parts are assembled and imported into

Hypermesh software. Meshing was created using

2D SHELL 63 elements to ensure connectivity and

manual adjustments. Then, 3D 10 Node structural

solid element SOLID 92 is used to create 3D

meshing of the whole assembly. There were seven

critical locations selected in the assembly. At these

locations 1D rigids were used. Spring seats center,

tie rod end center and wheel centers constitute the

said critical locations.

Fig. 1 Existing Model and sub-parts

36



Fig. 2 Concept Development Processes

Table.1 Experiment models

Parameters Existing

Shape Rectangular

Thickness 14

Material IS2062 IS2062 BSK 46 IS2062 BSK 46 IS2062 BSK 46 IS2062 BSK 46

IS 2062 : Yield Strength 350 N/mm2

BSK 46 : Yield Strength 460 N/mm2

Experiment Models

Square Circular

12.5 14 12.5 14

After creating tetra mesh, the 2D shell elements

were deleted. The rest were exported as FE model

to be used in ANSYS finite element solver. In the

existing model, the number of elements is 233372,

the number of nodes is 393582 and ‗Solid 92‘

elements were used. In all the experiment models,

the number of elements was controlled so as to

keep the number of elements less than 150,000 due

to computer system limitations and processing

duration.

NUMERICAL SIMULATIONS

The meshed models were imported into

commercially available ANSYS finite element

solver. This is an easy and powerful tool to get

numerical approximations with reasonable accuracy

and great speed. For the analysis four vital load

cases were considered. They are vertical or bending

which is static in nature. Other load cases such as

Acceleration, Braking and Cornering are dynamic

in nature. In order to simplify the problem a quasi-

static approach was used in the numerical

simulations. The job required to find the reactions

at the critical locations while the vehicle is

operating under extreme dynamic maneuvers in the

Acceleration, Braking and Cornering load cases.

This was done using ADAMS software on the

complete vehicle model and applying relevant loads

at the vehicle CG location. From this, the reactional

forces at the identified critical locations are

recorded.

After getting the reaction forces in the critical

locations, the same load vectors were applied using

inertia relief method in the same critical locations

in ANSYS finite element solver. The same

methodology was applied to all models. In this

method no constraints were used and only forces

applied to get the stress plots. This approach

simplifies the problem solving from dynamic or

transient load conditions into simple static load

37



condition. However, vertical load case was solved

as a simple static load condition.

Solving this and post-processing we can read the

contour nodal solution data for displacement

components and von-mises stress plots. For metals,

the Von Mises‘ yield function is a good description

while analyzing ductile metals. It gives valuable

inputs like the stress distribution, maximum stress

and location in graphical or table form.

The complete study and project work is vast that

the models, mesh and stress plots of all models

cannot be represented in this document. So, only

sample pictures are given for better understanding.

Given below are few sample pictures of Circular 14

model, showing the Von-mises stress distributions

when subjected to different load cases as shown

below.

Fig.3 Vertical or Bending Load case

Fig. 4 Meshing of Circular 14mm wall thick model

38



Fig.5 Acceleration Load case

Fig.6 Braking Load case Fig.7 Cornering Load case

Table 2 Induced Stress Levels (Von-mises Stress)

Fig.8 Vertical Load cases chart Fig.9 Acceleration Load cases chart

Induced Stress in N/mm2

Existing S12 S14 C14 C16

Vertical load case 124.4 159.2 138.2 169.6 156.9

Acceleration load case 297.8 324.5 263.7 276 272.2

Braking load case 159.7 155.5 155.5 133.3 139.2

Cornering load case 194.4 233.3 194.4 186.3 194.4

Vertical load case- Deflection in mm 1.238 1.768 1.619 3.542 1.844

0

50

100

150

200

250

300

350

Exi

stin

g

S1

2

S1

4

C1

4

C1

6

Vertical load case

Vertical load case

0

50

100

150

200

250

300

350

Exis

tin

g

S12

S14

C14

C16

Acceleration load case

Acceleration load case

39



Fig.10 Braking Load cases chart Fig.11 Cornering Load cases chart

Table 3 Weight Reduction

Weight Reduction Existing S12 S14 C14 C16

Model Weight in kgf 179.5 152.6 160.3 152.9 161.9

weight reduction in kgf 0 26.9 19.2 26.6 17.6

The Von-mises stress levels given above were almost

within the yield strength limit or at par with the

reference model stress levels. Also, the results of

using alternate material and reducing weight was

unsafe as seen in C12 model (So this was not

considered). With this, the scope for the usage of

alternate material depends if more factor of safety is

demanded.

The graphical representation of analysis results of the

experimental models is shown. From this we can find

out that the Acceleration load case is the severest

among the load cases. This is because load transfer

occurs from the front axle to the rear axle during

acceleration load case and vice versa in the braking

load condition.

The deflection in mm is found as 1.238 in the

existing model. In the experiment models the

deflection values in mm are found as 1.619 (S14),

1.768 (S12), 1.844(C16) and 3.542(C14). Though the

deflection found was 3.5 mm in C14, the same does

not affect the assembly in fitment and function.

The complete summary of analysis is tabulated as

below. The existing model values are kept for

reference. The same were plotted in graphical form

also for easy understanding.

The Von-mises stress levels given above were almost

within the yield strength limit or at par with the

reference model stress levels. Also, the results of

using alternate material and reducing weight was

unsafe as seen in C12 model.

From the graphical representation of analysis results,

it can be seen that the acceleration load case is the

most severe among different load cases. This is

because load transfer occurs from the front axle to the

rear axle during acceleration and vice versa in the

braking.

The deflection in the existing model is found to be

1.238 mm. In the experiment models the deflection

values in mm are found to be 1.619 (S14), 1.768

(S12), 1.844(C16) and 3.542(C14). Though the

deflection found was 3.5 mm in C14, the same does

not affect the assembly in fitment and function.

DESIGN OF EXPERIMENTS

Introduction

The weight reduction of the rear axle housing is the

objective. The DOE has three factors shape, thick and

material at two levels of experiment. However since

the material is not an input variable in the analysis,

let us consider the problem as 22 factorial design

leaving the material variable.

050

100150200250300350

Exis

tin

g

S12

S14

C14

C16

Braking load case

Braking load case

0

50

100

150

200

250

300

350

Exi

sting

S12

S14

C14

C16

Cornering load case

Cornering load case

40



Goal of experiments

• Experiments help us in understanding the

behaviour of a (mechanical) system

• Data collected by systematic variation of

influencing factors helps us to quantitatively

describe the underlying phenomenon

Full factorial design

A full factorial design of experiments consists of the

following:

- Vary one factor at a time

- Perform experiments for all levels of all

factors

- Hence perform a large number of

experiments that are needed

- All interactions are captured (as will be

shown later)

The two level values are assigned +1 and -1 as

explained in the case of 22

factorial experiments. A 22

factorial experiment was performed and the

following matrix gives the results.

The P is the parameter that is determined by using the

outcome matrix by the simultaneous solution of the

following four equations:

P0 + PA + PB + PAB = 276 (i)

P0 - PA + PB - PAB = 272.2 (ii)

P0 - PA - PB - PAB = 324.5 (iii)

P0 - PA - PB + PAB = 263.7 (iv)

From the above calculating the values dividing by 4,

as below:

(i)+ (ii) + (iii) + (iv):

276+272.2+324.5+263.7= 1136.4/4 = 284.1

(i)- (ii) + (iii)-(iv):

276-272.2+324.5-263.7= 64.6/4 = 16.15

(i)+ (ii)-(iii)-(iv):

276+272.2-324.5-263.7= -40/4 = -10

(i)- (ii)-(iii) + (iv)

276-272.2-324.5+263.7= -57/4 = -14.25

A simple regression model that may be used can have

up to four parameters. Thus we may represent the

regression equation as:

y = P0 + PA XA+ PB XB+ PAB XAXB

y = 284.1 +16.15 XA -10 XB -14.25 XAXB

The deviation with respect to the mean is obviously

given by

d = y - 284.1= 16.15 XA -10 XB – 14.25 XAXB

It may be verified that the total sum of squares (SST)

of the deviations is given by

SST = 4 * (P2A XA+ P

2B XB+ P

2AB XAXB) = 4 *

(16.152 + 10

2 + 14.25

2)

SST = 4 * (260.8 + 100 + 203.1) = 2255.5

The sample variance is thus given by

Contributions to the sample variance are given by 4

times the square of the respective parameter and

hence we also have,

SSA = 4 * 260.8 = 1043.2

SSB = 4* 100 = 400

SSAB = 4 * 203.1 = 812.3

Here SSA means the sum of squares due to variation

in level of XA and so on. The relative contributions to

the sample variance are represented as percentage

contributions in the following table 4 as below:

Table 4 Percentage contribution

Thus the dominant factor is the shape factor followed

by the interaction and then thickness. In this example

all these have significant effects and hence a full

factorial experiment is justified.

XA XB +1 -1

+1 276 272.2

-1 324.5 263.7

contribution % contribution

SST 2255.5 100.0

SSA 1043.2 46.3

SSB 400 17.7

SSAB 954.8 36.0

41



SUMMARY AND CONCLUSION

1. In the Existing model, induced stress in N/mm2

found in Vertical Load case as 124.4 N/mm2,

Acceleration Load case 297.8 N/mm2, Braking Load

case 159.7 N/mm2

and Cornering Load case 194.4

N/mm2. Yield strength of material is 350 N/mm

2.

This is reckoned as reference for the experiment

models.

2. Static analysis approximating the dynamic load

cases were carried out on the experiment models of

Circular 14, Circular 16, Square 12 and Square 14.

Out of these Square 14 and Circular 14 are found

better.

3. While, the induced stresses in Square 14 and in

Circular 14 are comparable, the weight reduction in

Circular 14 (26.6 kg) is better than Square 14 (19.2

kg).

4. The study concludes that the circular cross-

sectional 14 mm wall thick model is the best model

out of the experimental models while referring with

to the existing model. This satisfies the objective of

weight reduction and still working safely under the

simulated load cases.

5. The numerical analysis is time consuming process

and so more samples could not be used for DOE

study or to conduct optimization study. However,

this offers future scope to continue this project work

if there will be good demand for optimization.

ACKNOWLEDGEMENTS

The authors thank for the continuous and valuable

support extended by the members belongs to Anna

University –AUFRG- CAD/CAM Centre, Chennai

and Department of Product Development, Technical

Center, Ashok Leyland, Chennai.

REFERENCES

1. Brown.J, Robertson.J & Serpento.S (2002),

‗Motor Vehicle Structures, SAE

International Publications,

2. Chinnaraj.K, M.Satya Prasad and

Lakshmana Rao.C, (2009) ‗Dynamic

Response Analysis of a Heavy Commercial

Vehicle subjected to extreme Road

operating Conditions‘, Journal of Physics,

Conference series 181, 012070.

3. Chinnaraj.K, M.Satya Prasad and

Lakshmana Rao.C, (2008), ‗Experimental

Analysis and Quasi-Static Numerical

Idealisation of Dynamic Stresses on a Heavy

Truck Chassis Frame Assembly‘, Applied

Mechanics and Materials Vol. 13-14, pp271-

280

4. Gillespie.T (1992), ‗Fundamentals of

Vehicle Dynamics‘, SAE International

Publications, Society of Automotive

Engineers, 400 Commonwealth drive,

Warrendale, PA 15096, USA

5. Ji-Xin Wang, Guo-Qiang Wang, Shi-Kui

Luo, Dec-Heng Zhou, (2004) ‗Static and

Dynamic Strength on Rear Axle of Small

Payload Off-highway Dump Trucks‘,

CADFEM User‘s Meeting and Conference,

Germany,

6. John Fenton (1999), ‗Advances in Vehicle

Design‘, Professional Engineering

Publishing Limited, London and Bury St

Edmunds, UK

7. Mike Blundell and Damien Harty (2004),

‗The Multi-Body Systems Approach to

Vehicle Dynamics‘, SAE International

Publications,

8. Venkateshan. S.P., ‗Design of Experiments‘,

IIT-Madras

CONTACTS

1 PG Student, CEG, Anna University, Chennai,

India

[email protected]

2 Lecturer, Department of Mechanical

Engineering, CEG, Anna University, Chennai,

India

[email protected]

3 Divisional Manager, Product Development,

Ashok Leyland, Chennai, India

[email protected]

42



N-2010-D14-14

Piston Ring Pack Optimization for a DI Diesel Engine by

Predictive Technique Approach and Experimental

Verification

C Bharathi, N K Cherian, S Ganesan, S Mohanraj, R Mahadevan

India Pistons Ltd., Chennai, India

ABSTRACT

Emission norms for internal combustion engines

are becoming increasingly stringent and engine

manufacturers need to focus more on reducing the

engine emissions like CO, HC, NOx and particulate

matter in line with the target norms. The reduction

of oil consumption & blow-by plays an important

role in achieving this objective. For this reason the

OEM‘s strongly demand piston and ring assemblies

which can offer low oil consumption and blow-by.

Therefore it becomes necessary to study the axial

and radial motions of the rings and inter ring gas

pressures, as they are considered to be important

factors affecting oil consumption and blow-by. The

development of ring pack for low oil consumption

needs a lot of theoretical iterations and

experimental verification which are very costly and

time consuming, as the process involves a number

of prototypes and dynamometer tests. In order to

minimize multiple design iterations and testing,

organizations tend to use engine performance

simulation tools. In this paper, the approach

towards optimizing oil consumption in a four

stroke DI diesel engine has been presented. Ricardo

RINGPAK software has been used for the

simulation. Thermal analysis has been carried out

using Ricardo PISDYN software for predicting the

temperature distribution of piston and liner

required as input for the RINGPAK analysis.

INTRODUCTION

The achievement of future engine development

need requires a better understanding of the

fundamentals of cylinder kit dynamics. The

cylinder kit dynamics affect Oil consumption,

emissions and friction and in the recent years,

analytical modelling is being extensively used for

this study. However, this new understanding can

only be fruitful if analytical predictions are

validated by experimental results. The operation of

rings in piston ring pack is based on the

interactions between various physical phenomena

such as ring axial and radial motions, ring twist,

gas flow through the end gap and land clearances,

ring bore conformability, transport of oil and

hydrodynamic lubrication.

Several studies have been conducted in the

modelling of these various coupled phenomena in

an integrated manner. Experimental investigation

for oil consumption and blow-by in relation to

piston and ring features was reported in an early

paper by N.A. Graham and R. Munro [1]. Inter-ring

gas pressures and blow-by in a diesel engine were

investigated analytically and compared with the

measured data by Zafer Dursunkaya, Rifat Keribar

[2]. Christopher G. Knowland and Christopher J.

Russell [3] have developed a numerical code to

study the piston and ring stability and their

influence on oil consumption and blow-by. D.E.

Richardson [4] tested two different ring pack

configurations with the second ring having positive

twist and negative twist and reported significant

difference in measured inter ring pressures. Hans

H. Priebsch and Hubert M. Herbst [6] have done

modelling of cylinder kit dynamics of a diesel

engine for several operating conditions and ring

modifications.

SCOPE OF WORK

In the present study there was a requirement to

reduce the oil consumption by 30 % in a 4 litre

naturally aspirated DI diesel engine based on

customer need. Tested values of Oil consumption

and blow-by with the base configuration of piston

43



and rings were 16 g/hr and 28 lpm respectively.

The simulation study was then conducted, firstly

for the base engine configuration, using the

simulation software, RINGPAK. Simulation results

of oil consumption and blow by were in line with

the tested values. Effect of top ring stability,

second ring closed gap, top ring barrel and oil ring

land width on oil consumption were studied

individually and in combination using RINGPAK

simulation. Based on the above mentioned studies

final combination of ring pack was suggested for

improvement of oil consumption. Suggested ring

pack was tested and found that oil consumption and

blow-by values are within the target values.

ENGINE SPECIFICATION AND BASE

PISTON & RING CONFIGURATION

The technical specification of the engine used for

the study is given in table 1 and the base

configuration of the piston and rings is given in Fig

1. The targeted Value of O/C is 10 g/hr.

Table1 Engine Specifications

Parameter Details

Type 4-Stroke, NA, DI

No of cylinder 4, Inline

Bore x Stroke (mm) 100 x 127

Cubic capacity(cc) 3987.8

Rated power (hp) 76

Peak pressure ( bar) 61 max

Rated Speed (rpm) 2200

Fig 1 Base Configuration of Rings

FINITE ELEMENT ANALYSIS

Piston temperature distribution was obtained by

performing Thermal analysis of piston using

analysis software, PISDYN. Area averaged & time

mean temperatures and convective heat transfer

coefficients of the cylinder gas were applied on the

piston crown. Gas convective heat transfer

coefficients and bulk temperatures were applied on

the various areas like lands, groove sides, boss, and

skirt and under side of the piston crown.

Fig 2 Piston Temperature Distribution

In a similar manner, thermal analysis of liner was

carried out and the temperature distribution of the

liner predicted. These predicted temperatures of the

piston and liner were used as one of the inputs for

RINGPAK analysis. The predicted piston

temperature distribution is shown in Fig 2.

RINGPAK SOFTWARE

Ricardo‘s RINGPAK software (version 5) was

used as a simulation tool for investigating the

lubricating oil consumption and ring dynamics in

the existing DI engine ring pack (base)

configuration.

INPUT DETAILS

1. Piston ring pack configuration and

geometry

2. Material properties of the components

3. Surface finish parameters of the

components

4. Lubricant properties

5. Engine Operating conditions like

a) Engine Speed

b) Cylinder gas temperature

44



c) Cylinder gas pressure

d) Crankcase gas temperature

e) Crankcase pressure

6. Temperature details of the

piston/ring/liner

7. Distorted bore shape

RINGPAK WORKING PRINICPLE

The RINGPAK module predicts oil consumption

and blow-by as an outcome of calculation of a set

of seven complex modules which are

interconnected. The simulation considers the

following three main mechanisms of oil loss from

the cylinder kit.

a) Oil consumption due to evaporation: - The oil

film distributed in the liner surface in the top

portion gets exposed to the combustion gases and

consequent high temperature. As the piston moves

down this oil film evaporates and causes oil loss.

b) Oil throw from the top ring: - During the

upward motion of the piston some oil is carried in

the leading edge of the top ring and at the time of

TDC reversal is thrown towards the combustion

chamber.

c) Oil carried by blow-back gases: - Some oil is

carried by the gasses that blow back into the

combustion chamber from the crevices in the

clearances between piston and bore as well as rings

and grooves.

Cumulative effect of all the three phenomena

results in an estimate of overall oil consumption.

OUTPUT DETAILS

RINGPAK software provides three different types

of output

1. Cycle averaged results related to ring pack

performance which include

a. Oil Consumption

b. Gas blow-by and blow-back

c. Friction and power losses

d. Wear rate of ring faces and groove-ring

side faces and liner

2. Animation files showing the gas pressures, gas

temperatures, gas mass flow rates through the

piston land and grooves for complete cycle of 720

deg crank angle.

3. Performance plots of the ring pack which include

a. Ring axial and radial motions

b. Groove and land pressures

c. Oil flow rates through the ring faces

d. Blow-by mass past top ring etc

RINGPAK MODELING

The following parameters were used for modelling:

Bore, stroke, connecting rod length, coordinates of

piston lands, grooves, rings and ring faces,

pressure-crank angle, temperature-crank angle,

surface roughness, surface hardness. Temperature

details of piston, rings and liner (from the FE

analysis results).

TOP RING STABILITY (OPTIMIZING

INTER RING PRESSURE)

Base configuration of the piston and ring pack was

analyzed for oil consumption and blow by. Initially

rings axial motions in the grooves were studied. It

was observed that the top ring was unstable.

Duration of the top ring lift is high, starting from

expansion stroke and ending at the middle of the

intake stroke as shown in Fig 3. It indicates that the

top ring stays for a greater duration near the top

groove top flank. As a result, the oil available at the

top edge of the top ring top is more prone to be

thrown into the combustion chamber. This higher

duration lift of top ring in a cycle could be due to

greater second land pressure in comparison to the

top land pressure, as highlighted in Fig 4. In order

to reduce the second land pressure for better land

pressure balancing, second ring end gap was

increased in steps of 0.10mm. At 0.85mm second

ring gap, second land pressure reduced and became

equal to the top land pressure as shown in Fig 5.

Because of this the top ring became more stable.

The duration of top ring lift also is reduced

considerably as shown in the Fig 6. This will have

a positive effect on oil consumption reduction.

45



Fig 3 Top Ring Axial Position in Groove

Fig 4 Top, 2nd

and 3rd

Land Pressures

Fig 5 Top and Second Land Pressures with Higher

Second Ring Gap

Fig 6 Top Ring Axial Position in Groove with

Higher Second Ring Gap

SECOND & OIL RING POSTIONS

In both base and modified condition, second and oil

rings are lifting twice in the groove, once in the

middle of compression stroke and once at the end

of exhaust stroke. Second and oil ring axial

motions in the groove for initial configuration are

shown in the Figs.7 and 8. Intentionally the axial

behaviour of the second and oil ring in the groove

is left as it is in order to study the effect of first ring

dynamics alone.

Fig 7 Second Ring Axial Position in Groove

Fig 8 Oil Ring Axial Position in Groove

EFFECT OF BARREL ON TOP RING

The base configuration top ring has symmetric

barrel on its OD profile. With the introduction of

an offset barrel, the point of contact of the ring face

46



with the bore shifts towards the bottom side of the

ring. This shift of the barrel contact point helps the

ring to glide better on the oil film during upwards

strokes there by reducing quantum of oil being

pushed up. Lowering the contact point also helps in

better scraping of the oil during the downward

strokes. Simulation was carried out to study the

effect of top ring barrel on oil consumption by

varying the barrel configuration as Offset barrel

and asymmetric barrel and retaining the other

parameters. From the simulation results, oil

accumulation at the top ring leading edge (moving

edge) is less during compression stroke (less throw)

and more during downward strokes (better

scraping). But in exhaust stroke, it is more or less

equal to the base case. With asymmetrical barrel,

oil accumulation in all strokes is satisfactory.

Comparisons of the oil accumulation trend for both

the cases along with base case are shown in the

Fig 9.

Fig 9 Comparison of Oil Accumulation on the Top

Ring leading edge

ROLE OF OIL RING LAND WIDTH

Reduction of oil ring land width results in higher

pressure being applied by the land for the same

tangential load. Simulation exercise was carried out

to study the effect of oil ring land width and it has

been observed that lower ring land width results in

better oil scarping. Change in the oil ring land

width from 0.35 to 0.25mm considerably increases

the scraping capability of the oil ring. Comparison

of the oil scraping capacity is shown in Fig.10.

Fig 10 Effect of oil ring land width in Oil

Scrapping

PROPOSED MODIFICATIONS

Based on the independent simulation studies,

namely top ring stability, second ring closed gap

changes, top ring barrel effect and oil ring land

width effect, the following modifications was

proposed in table 2.

Using this proposed combination, RINGPAK

simulation was carried out and the following were

studied

(a) Top ring stability

(b) Oil accumulation on the top ring leading edge

(c) Oil Consumption and Blow by

Table 2 Proposed Modifications

Ring Parameter Base Proposed

Top

ring Barrel Symmetric Asymmetric

Second

ring

Closed

gap(mm) 0.55 0.85

Oil ring Land width

(mm) 0.35 0.25

It has been observed that top ring was relatively

much stable in comparison to the base

configuration. Comparison of top ring axial

position in the groove for base and proposed design

is shown in Fig 11. Oil accumulation at the top ring

leading edge was reduced during upward strokes.

Comparison of oil accumulation at the top ring

leading edge for base and proposed design is

shown in Fig 12. A simulated result shows very

good improvement in oil consumption. Comparison

47



of the simulated values of oil consumption and

blow-by for both base and proposed configurations

is shown in Fig.13.

Fig 11 Comparison of Top Ring Axial Position in

the Groove for Base and Proposed Designs

Fig 12 Comparison of Oil Accumulation at the Top

Ring Leading Edge for Base and Proposed

configurations

Fig 13 Comparison of O/C and BBY for base and

proposed design

EXPERIMENTAL VERIFICATION

The engine was tested with the proposed ring pack

and Oil consumption test was carried out at rated

speed and full load. Oil consumption value is

observed to be within the target value at 10 g/hr.

Blow by value recorded was 27 lpm against the

base value of 28 lpm. Comparison of experimental

values of oil consumption and blow-by for base

and proposed design are shown in Fig.14.a and

14.b respectively. Comparison of simulated and

experimental values of oil consumption for

proposed design is shown in Fig. 15.

Fig 14.a Comparison of Experimental Values of

O/C for Base and Proposed Design

Fig 14.b Comparison of Experimental Values of

BBY for Base and Proposed Design

O/C Comparison ( g/hr )

Base config Proposed

config

Prediction 19 14

Experimental 16 10

% Variation 16 29

Fig 15 Comparison of Simulated and Experimental

Values of O/C

0.00

5.00

10.00

15.00

20.00

25.00

30.00

Base Proposed

OC (g/hr)

Bby(lpm)

Oil Consumption (g/hr)

0.00

5.00

10.00

15.00

20.00

Base Proposed

Simulation

Experimental

Blow by (lpm)

0.00

5.00

10.00

15.00

20.00

25.00

30.00

Base Proposed

Simulation

Experimental

48



CONCLUSIONS

The target values of oil consumption and blow by

was achieved with the proposed ring pack using

RINGPAK simulation

Increase in the second ring end gap

allows easier flow of gases and oil to the

third land, thereby reducing the second

land gas pressure.

Reduction in the second land gas pressure

makes the top ring highly stable. The

duration of top ring lift is drastically

reduced. This in turn reduces the oil

available for throw to the combustion

chamber leading to lower oil throw-off.

Reduction in second land gas pressure

also reduces blow-back through the

closed gap of the top ring, resulting in

lower oil consumption.

Asymmetric Barrel profile in top ring

periphery facilitates better gliding of the

ring on the oil film. This improves the

scraping capacity of the ring.

Asymmetric barrel profile in top ring

causes less oil accumulation at the top

ring leading edge (better gliding) during

upward strokes and effective oil scraping

in the downward strokes.

Reduction of oil ring land width improves

the oil scraping efficiency.

RINGPAK simulation is a helpful tool to predict

relative oil consumption and blow by levels and

reduces the number of theoretical iterations and

expensive testing.

ACKNOWLEDGMENT

The authors wish to thank the management of India

Pistons Ltd for providing the necessary support for

carrying out this study.

REFERENCES

1. N. A. Graham, R. Munro, (1979)

‗Investigation and Analysis of Oil

Consumption and Blow-By in Relation to

Piston and Ring Features‘, AE.

Symposium, paper No 28. Pp.1-9.

2. Zafer Dursunkaya, Rafit Keirbar, Dana

E.Richardson., (1993) ―Numerical and

Experimental Investigation of Inter-Ring

Pressures and Blow-by in a Diesel

Engine‖. SAE-930792., pp.1-9

3. Christopher G. Knowland, Christopher J.

Russell., (1996) ‗Predictive Optimization

of Piston and Ring Stability‘. SAE-

960873., pp.225-230

4. D. E. Richardson, (1996) ‗Comparison of

Measured and Theoretical Inter-Ring Gas

Pressure on a Diesel Engine‘. SAE-

961909, pp.1910-1923

5. Jinglei Chen, D. E.Richardson, (1999)

‗Predicted and Measured Ring pack

Performance of a Diesel Engine‘, 4th

Ricardo user‘s conference.

6. Hubert M. Herbst, Hans H. Priebsch.

(2000) ‗Simulation of Piston Ring

Dynamics and Their Effect on Oil

Consumption‘, SAE-2000-01-0919.,

pp.862-873

7. Ricardo‘s RINGPAK user‘s manual,

version 5

49


Rajalakshmi Engineering College, Thandalam, Chennai: 15th

& 16th

July 2010

N-2010-E04-16

Effect of EGR and DPF on Emission of DI Diesel Engines

to Meet BS IV Norms

1Sundararaman.R,

2Pugazhvadivu.M,

3Sankaranarayanan.G,

4Jeyachandran.K

Keywords: IC engine, Emission, EGR, DPF

ABSTRACT

Surface transport is the backbone of a nation which

keeps the day to day activities of a country live.

Worldwide pollution by transport systems is a

major ecological threat. In general, most of the

surface transport vehicles use diesel powered

engines. Diesel engines offer higher thermal

efficiency and durability. They are mostly emitting

very low CO and HC due to lean operations at the

same time at higher loads they emit particulates

too. In diesel engines, NOx formation is a

temperature-dependent phenomenon and takes

place when the temperature in the combustion

chamber exceeds 2000 K. Therefore, in order to

reduce NOx emissions in the exhaust, it is

necessary to reduce peak combustion temperatures.

One such way of reducing the NOx emission of a

diesel engine is exhaust gas recirculation (EGR).

Re-circulating exhaust gas helps in reducing NOx,

but appreciable increase in particulate emissions

are observed at high loads, hence there is a trade-

off between NOx and smoke emission. To get

maximum benefit from this trade-off, a particulate

trap was used to reduce the amount of unburnt

particulates in EGR, which in turn reduces the

particulate emission also. The present investigation

is to study the effects of hot and cooled EGR on

emission characteristics with and without

particulate trap. Experiments were conducted for

observing the effect of different quantities of EGR.

It can be observed that the combined operation of

EGR and DPF showed that the NOx reduced by

63% and the smoke by 20% although the SFC

increased by 13%.

INTRODUCTION

In the recent years, stringent emission legislations

have been imposed on NOx, smoke and particulate

emissions from automotive diesel engines. Diesel

engines are typically characterised by low fuel

consumption and very low CO emissions.

However, the NOx emissions from diesel engines

still remain high. Hence, in order to meet the

environmental legislations, it is highly desirable to

reduce the amount of NOx in the exhaust gas.

Diesel engines are predominantly used to drive

tractors, heavy vehicles and trucks. Owing to their

low fuel consumption, they have become

increasingly attractive for smaller trucks and

passenger cars also. But higher NOx emissions

from diesel engine remains a major problem in the

pollution aspect. In the present work, an attempt is

made to control the engine exhaust emissions using

EGR combined with particulate trap fitted in a

diesel engine.

Exhaust gas recirculation (EGR) has been used in

recent years to reduce NOx emissions in diesel

engines. EGR involves diverting a fraction of the

exhaust gas into the intake manifold where the

recirculated exhaust gas mixes with the incoming

air before being inducted into the combustion

chamber. EGR reduces NOx because it dilutes the

intake charge and lowers the combustion

temperature. At high loads, EGR suppresses flame

speed sufficiently that combustion becomes

incomplete and unacceptable levels of particulate

matter (PM) and hydrocarbons (HC) are released in

the exhaust.

50



EXPERIMENTAL SETUP

The experiments were conducted on a single

cylinder, naturally aspirated, air-cooled, Kirloskar

DI Diesel engine coupled to an electrical generator.

The specifications of the engine are given in Table

1. The engine was run at rated speed of 1500 rpm.

The performance of the engine was evaluated in

terms of SFC, brake thermal efficiency, and

emission characteristics like HC, CO, NOx, and

smoke were recorded for cooled EGR, hot EGR

and particulate trap. Further the combined effects

of cooled EGR and DPF (Diesel particulate Filter

also known as Particulate trap) were also

investigated.

EGR pipe was connected from the outlet of the

exhaust manifold to the inlet of the intake

manifold. EGR rate was regulated by controlling

the valve-A, which is installed in the intake

manifold, the junction between the EGR pipe, and

the intake manifold inlet is plugged when EGR is

inactive. The EGR percentages were varied at

different rates in steps of 5 to a maximum of 50 %.

The catalyst used in this study was bimetal

catalysts namely copper oxide and zinc oxide. The

supported catalyst was a pellet made of silica

(clay). The catalyst was packed inside the converter

and it was fitted in the exhaust pipe

The quantity of EGR was measured and controlled

accurately; hence a by-pass for the exhaust gas was

provided along with the manually controlled EGR

valve. The exhaust gas comes out of the engine

during the exhaust stroke at high pressure. It is

pulsating in nature. It is desirable to remove these

pulses in order to make the volumetric flow rate

measurements of the recirculating gas possible.

For this purpose, another smaller air box with a

diaphragm is installed in the EGR route. An orifice

meter was installed to measure the volumetric flow

rate of the EGR.

EGR rate was defined as follows:

In this study, EGR rate was determined as the ratio

of CO2 concentration in intake gases to that in

exhaust gases, because .

Therefore

Thermo-couples are provided at the intake

manifold, exhaust manifold and various points

along the EGR route to measure the temperature.

For cooled EGR, water-cooled recirculated gas was

directly induced in the intake-pipe and the gas flow

was regulated with an EGR control valve. Inlet gas

temperature can be kept close to room temperature

by using EGR cooler.

The temperatures of the intake gas, exhaust gas and

inlet/outlet cooling water were measured using

thermocouples. When EGR cooler was not used,

the temperature of intake gas varied EGR gas flow

rate.

For cooling 16-tube shell and tube type cross flow

heat exchanger was used. In hot EGR, the exhaust

gas is diverted into the intake manifold without

cooling. In order to avoid heat loss, the pipelines

were made short and insulated.

One of the main parameter to take into account

was the distance between the EGR induction point

and the intake manifold. The catalytic converter

attachment was connected to the engine exhaust

line by a pipe such that the exhaust gas enters the

converter axially. The catalytic converter was

attached close to the engine exhaust outlet pipe.

Silica pellets (clay) were packed in a cylindrical

container and both ends were fixed with stainless

steel wire mesh.

Fig. 1 Experimental Setup

51



Nearer to the catalytic converter a pocket was

provided to measure the bed temperature using

thermocouple. Every experiment was repeated to

ensure consistency of the results.

Table 1. Specification of the test engine

Make & Model Kirloskar,single Cyl, D.I

diesel engine Vertical

Bore & Stroke 87.5 mm x 110 mm

Piston

displacement 661 cc

Output 4.41kW @ 1500 rpm

Comp. ratio 17.5:1

Cooling Air cooled

Fuel spill time 23 BTDC


Fig.2 shows the variation of brake thermal

efficiency with respect to brake power under

various EGR rates. The figure shows that the brake

thermal efficiency decreases with an increasing

EGR rate.

Fig.2 Variation of Brake thermal efficiency

At full load the thermal efficiency decreased slowly

up to 15 % EGR and then decreased drastically.

The reduction in thermal efficiency with 15%

cooled EGR is 6% and when the engine fitted with

EGR and particulate trap leads to reduction by

about 11% at maximum load.

The reason may be due to the replacement of

oxygen below 17% in the intake air causes drastic

reduction in combustion temperature, which leads

to reduction in combustion efficiency.

Fig.3 shows the variation of brake specific fuel

consumption (BSFC) with respect to brake power

under no load to full load at various EGR rate. At

15% EGR the SFC increased to 0.34 kg/kWh i.e.

about 9% increases in fuel consumption. SFC was

further increased when the engine fitted with DPF.

Specific fuel consumption for the engine, when

EGR combined with trap increases by about a

maximum of 13%. The reason was, due to increase

in backpressure.

The increase in EGR rate shows that there was a

slight increase in BSFC due to the reduction in

oxygen content in the intake, which resulted in

reduction in combustion temperature.

Fig. 3 Variation of S.F.C

The increase in EGR rate shows that there was a

slight increase in BSFC due to the reduction in

oxygen content in the intake, which resulted in

reduction in combustion temperature.

Fig.4 shows the variation of oxides of nitrogen with

brake power. The NOx decreases rapidly with

0

5

10

15

20

25

30

0 2 4

BR

AK

E T

HE

RM

AL

EF

FIC

IEN

CY

, %

BRAKE POWER, kW

0%10%15%20%30%40%50%

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 2 4

S.F

.C, kg/kW

.h

BRAKE POWER, kW

0%

10%

15%

20%

30%

40%

50%

52



increasing EGR. The NOx concentration increases

exponentially with increasing intake O2. The NOx

concentration was very low at O2 concentration

under 14%. A slight decrease of O2 starting from

the atmospheric concentration 20.8% leads to

strong decrease of NOx. The NOx reduction

becomes less pronounced at O2 concentrations

below 17%, Engine operation less than 16% O2 is

not realistic since the reduction in intake oxygen

leads to reduce the flame temperature with in turn

reducing the formation of NOx and fuel does not

burn completely and HC and particulate emissions

increase dramatically.

Fig. 4 Variation of NOx emissions

EGR significantly reduces NOx more at high loads

than at low loads, because at high loads oxygen

concentration in exhaust gases was lower which in

turn results in lower intake gas oxygen

concentration at the same EGR rate. The variation

of intake O2 shows a great potential for the

reduction of NOx emissions.

Fig.7 shows the difference of hot EGR (natural

cooling) and cool EGR (cooling by EGR cooler) on

NOx reduction under different EGR rate. The

temperature indicated at each point was measured

in the intake manifold, after mixing with EGR gas.

For both cool and hot EGR, NOx reduction has a

strong correlation with oxygen correlation in the

intake gas. Fig.8 shows that intake gas

temperatures also affect NOx reduction. At a given

level of oxygen concentration, the cool EGR

reduces more NOx which in turn indicates that, to

achieve certain NOx reduction, cool EGR can

achieve the goal with less EGR rate than does hot

EGR. The NOx emission was nearly zero at EGR

rates over 50%. More than 60% EGR was

technically not useful because of high HC and

particulate emissions. At 50% EGR flame

quenching leads to raising HC emissions. The

reduction of NOx for 15% EGR is 57% with cooled

EGR and 45% with hot EGR. When the engine was

fitted with trap it‘s around 62% with cooled EGR.

Fig. 5 Variation of smoke emissions

The results show that, only a marginal variation in

UBHC emission was noticed up to 15% EGR and

then increased drastically. The reason for this

phenomenon was the replacements of oxygen in the

inlet charge resulted in reduced oxidation and

lower gas temperature during expansion and

exhaust process and further the reduction of O2

below17% in the intake air causes drastic reduction

in combustion temperature which increases the HC

level drastically. It can also be seen that

combustion degradation dramatically increases CO

and HC levels above 15% EGR. At medium load,

characterized by higher overall A/F levels,

demonstrate almost no change in HC or CO levels

with EGR. At full load condition HC emission in

15% EGR increased to 25%. The reason was less

air entered resulting in inefficient combustion. The

CO emission of the diesel was 0.63 % vol at full

load and 1.2% vol for 15% EGR.

0

200

400

600

800

1000

1200

1400

1600

1800

0 1 2 3 4 5

OX

IDE

S O

F N

ITR

OG

EN

, p

pm

BRAKE POWER, kW

0%

10%

15%

20%

30%

40%

50%

0

1

2

3

4

5

6

7

0 2 4

SM

OK

E I

NT

EN

SIT

Y (

Bo

sch

)

BRAKE POWER,kW

0%

10%

15%

20%

30%

40%

50%

53



The variation of smoke with power output is shown

in fig.5. Smoke intensity under a constant load

tends to increase with the increase of EGR rate.

This tendency was considered to depend on the

reduction of oxygen in the incoming charge due to

EGR application. However at a given NOx level

the cool EGR results in lower smoke emission than

hot EGR.

Fig. 6 Variation of NOx reduction rate with

EGR rate

On the other hand, smoke emission increases partly

because the duration of diffusion combustion was

extended by EGR. Since EGR causes premix

combustion to slow down the fuel, which does not

burn in the premix combustion, burns later,

resulting in longer duration. Large amount of soot

that remains un-oxidised during the diffusion

combustion because of EGR

Smoke emission depends on the absolute amount of

oxygen in intake gases regardless of EGR types.

Smoke emissions increased with EGR rate until

approximately 40% .At this point a further increase

in EGR rate resulted in a decrease in smoke

emissions, since the colour of the exhaust slowly

changed from black to grey and then to white. The

smoke concentration increased by about 30% when

it was operated with 15% EGR. Further, when the

engine was fitted with DPF, there is no significant

difference in the smoke level at zero loads. This is

due to the lower trap efficiency of the DPF at low

loads but as the load was increased DPF efficiency

also increases. At the maximum load, a maximum

reduction of 20% is achieved from base mode when

the engine was operated with EGR and particulate

trap.

Fig. 7 Variation of inlet temperature with

EGR rate

Fig. 8 Variation of NOx emissions with EGR rate

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0% 20% 40% 60%

NO

x R

ED

UC

TIO

N %

EGR RATE, %

0

25%

50%

75%

100% 20

30

40

50

60

70

80

0% 5% 10% 15% 20% 25%

INL

ET

TE

MP

ER

AT

UR

E ,

'C

EGR RATE,%

100%Load Cold EGR

100%Load HOT EGR

0

200

400

600

800

1000

1200

1400

1600

1800

0% 5% 10% 15% 20%

NO

x, p

pm

EGR RATE %

FULL LOAD COLD EGR

FULL LOAD HOT EGR

54



Fig. 9 Variation of smoke emissions with

EGR rate

Fig. 10 Variation of Brake thermal efficiency with

Load

Fig. 11 Variation of S.F.C with Load

Fig. 12 Variation of NOx emissions with Load

3

3.5

4

4.5

5

5.5

6

6.5

0% 5% 10% 15% 20%

SM

OK

E, B

osch

EGR RATE %

100%LOAD COLD EGR

100%LOAD HOT EGR

0

5

10

15

20

25

30

25% 50% 75% 100%

B.T

.E %

LOAD IN %

BASE 15% COOLED EGR

15% EGR+TRAP

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

25% 50% 75% 100%

S. F

. C

, kg

/kW

.h

LOAD IN %

BASE COOLED EGR EGR+TRAP

0

200

400

600

800

1000

1200

1400

1600

1800

0% 25% 50% 75% 100%

NO

x,p

pm

LOAD IN %

BASE 15% COOLED EGR 15%EGR+TRAP

55



Fig. 13 Variation of smoke emissions with Load

CONCLUSION

An experimental set-up to measure the effects of

exhaust gas recirculation on engine characteristics

like exhaust gas temperature, thermal efficiency,

brake specific fuel consumption and smoke opacity

was developed. Experiments were carried out using

the setup to prove the efficiency of EGR as a

technique for NOx reduction.

Exhaust gas temperature is reduced

drastically, by employing EGR. This

indirectly shows the potential for reduction of

NOx emission. This can be concluded from

the fact that the most important reason for the

formation of NOx in the combustion chamber

is the high temperature of about 2000 K at the

site of combustion.

Thermal efficiency and brake specific fuel

consumption are not affected significantly by

EGR.

Particulate matter emission in the exhaust

increases, as evident from smoke opacity

observations

Increase in particulate matter emissions due to

EGR can be taken care by DPF and adequate

regeneration techniques.

REFERENCES

1. Arcoumanis.C, Nagwaney.A, Hentschel.W,

Ropke.S, ― Effects of EGR on spray

development, combustion and emissions in a

1.9L direct injection diesel engine‖, SAE

952356.

2. David.L.Mitchell, John A.Pinson, Thomas

A.Litzinger, ―The effects of simulated EGR

via intake air dilution on combustion in an

optically accessible D.I. diesel engine‖, SAE

932798.

3. Dimitrios Psaras and Jerry C.Summers,

―Achieving the 2004 heavy duty diesel

emissions using electronic EGR and a cerium

based fuel borne catalyst‖, SAE 970189

4. Durnholz.M, Eifler.G, Endres.H, ―Exhaust gas

recirculation-A measure to reduce exhaust

emissions of diesel engines‖, SAE 920725

5. Emig.B, Gmehling.B, Popovska.N ,

Holemann.K, ―Passive regeneration of

catalyst coated knitted fibre diesel particulate

traps‖, SAE 960138.

6. Jenkin.M, Kawanami.M, Horiuchi.M,

Klein.H, ―Development of oxidation and De-

NOx catalyst for high temperature exhaust

diesel trucks‖, SAE 981196.

7. Murphy.M.J, Hillenbrand.L.J and

Trayser.D.A, Wasser.J.H, ―Assesment of

direct pariculate control –direct and catalytic

oxidation‖, SAE 810112.

8. Noboru uchida, Yasuhiro daisho, Takeshi

sailo, Hideaki sugano, ―Combined effects of

EGR and supercharging on diesel combustion

and emissions‖, SAE 930601,

9. Robert c.Yu, Syed m.shahed, ―Effects of

injection timing and exhaust gas recirculation

on emissions from a D.I.diesel engine‖, SAE

811234 p.no 1-11.

10. Ropke.S, Schweimer.G.W, Strauss.T.S, ―NOx

formation in diesel engines for various fuel

and intake gases‖, SAE 950213.

ABBREVIATIONS

1. D.I: Direct Injection

2. EGR: Exhaust Gas Recirculation

3. DPF: Diesel Particulate Filter

4. SFC: Specific Fuel consumption

5. CO: Carbon monoxide

6. CO2: Carbon dioxide

7. NOx: Oxides of Nitrogen

8. BSU: Bosch Smoke Unit

CONTACTS

1Sr.Lecturer, Indian Naval Academy, Ezhimala,

Kerala

2APME Pondichery Engg College, Pondichery

3PME. Adhiparasakthi Engg College,

Melmaruvathur

4Principal, SREC, Vandalur

0

1

2

3

4

5

6

0 25% 50% 75% 100%

SM

OK

E , B

SU

LOAD IN %

BASE

15% COOLED EGR

15% EGR+TRAP

56



N-2010-D15-17

Utilization of vehicle frontal pressure to improve engine’s

volumetric efficiency

1P . Myilavan, K . Arunachalam, M. M. Murali Suresh

Dept of Automobile Engg, MIT Campus, Anna University-Chennai

Keywords: IC Engine, emission, volumetric efficiency, air intake.

ABSTRACT

In this work, the velocity of the air acting in front

of the vehicle is utilised to improve the engine‘s

volumetric efficiency. This is achieved by placing a

diffuser appropriately such that the incoming air is

pressurised and well directed towards the intake

manifold by placing a duct suitably.

This modification considerably improves the

volumetric efficiency of the engine. Also CO, HC

emission will be reduced. However, there will be a

little increase of NOx emission due to increase in

peak temperature.

The diffuser arrangement reduces the frontal

projected area and hence reduces aerodynamic

drag, which in turn further improves the vehicle

performance.

INTRODUCTION

Volumetric efficiency is a ratio (or percentage) of

what volume of fuel and air actually enters the

cylinder during induction to the actual capacity of

the cylinder under static conditions. Therefore,

those engines that can create higher induction

manifold pressures (above ambient) will have

greater efficiencies. Volumetric efficiencies can be

improved in a number of ways, but most notably

the size of the valve openings compared to the

volume of the cylinder and streamlining the ports.

Engines with higher volumetric efficiency will

generally be able to run at higher speeds

(commonly measured in RPM) and produce more

overall power due to less power loss in moving air

in and out of the engine.

A common approach for manufacturers is to use

larger valves or multiple valves. Larger valves

increase flow but weigh more. Multi-valve engines

combine two or more smaller valves with areas

greater than a single, large valve while having less

weight. Carefully streamlining the ports increases

flow capability. This is referred to as Porting and is

done with the aid of an air flow bench for testing.

Many high performance cars use carefully arranged

air intakes and tuned exhaust systems to push air

into and out of the cylinders, making use of the

resonance of the system. A more modern

technique, variable valve timing, and attempts to

address changes in volumetric efficiency with

changes in speed of the engine: at higher speeds the

engine needs the valves open for a greater

percentage of the cycle time to move the charge in

and out of the engine.

Volumetric efficiencies above 100% can be

reached by using forced induction such as

supercharging or turbo charging. With proper

tuning, volumetric efficiencies of naturally-

aspirated engines can be improved. These engines

are typically of a DOHC layout with four valves

per cylinder.

SCHEMATIC LAYOUT OF THE PROJECT

In this project, air flow is guided to the engine by

means of a diffuser with a duct to the intake

manifold. The diffuser is placed in front of the

vehicle. The diffuser converts the high velocity air

to pressurised air. The length of the duct is used to

minimize the back pressure flow. When the vehicle

moves faster and faster the air flow increases and

compensates the reduction in volumetric efficiency

loss at high engine RPM and also improves the

performance of the vehicle.

57



Fig1. Schematic layout of the project

VEHICLE PERFORMANCE TESTS IN

CHASSIS DYNAMOMETER

Chassis dynamometer is a special type of

dynamometer, where an entire vehicle can be tested

for its performance. In chassis dynamometer the

vehicle is kept in a simulated road condition. With

chassis dynamometer the vehicle can be given load,

for which its power developed, acceleration, can be

obtained digitally in the computer. Chassis

dynamometer consists of motor/generator

connected to rollers, control panel, output panel,

etc.

Here the different air velocity is provided by means

of a constant discharge blower connected with an

setup with adjusting flap to vary the amount of air

flowing.

Fig 2 Chassis Dynamometer Control Unit

EXPERIMENTAL PROCEDURE

Vehicle performance test is conducted using

chassis dynamometer. The vehicle‘s driving wheels

are mounted in the dynamometer rollers. The load

is given as the opposing tractive force in the rollers

by giving resistance to the generator of the

dynamometer.

The fuel supply line for the vehicle is modified to

measure the fuel consumption. The intake duct is

connected to the surge tank with U-tube manometer

to measure the volumetric efficiency.

The loads given are 100, 150, for which the fuel

consumption and volumetric efficiency are

measured for the vehicle at 20, 30, 40, 50, 60 and

70 kmph.

Driving cycle followed for the vehicle testing in the

chassis dynamometer.

20 kmph - 2nd gear

30 kmph - 3rdgear

40 kmph - 4thgear

50 kmph and above -5thgear

Since the vehicle is in stationery position, for

different vehicle speed the air is fed to the manifold

from the blower through the valve setup. The valve

set up is adjusted for different vehicle velocity to

obtain the calculated air flow.

Fig 3 Valve setup connected with blower

58



Fig 4 Flap adjustment in valve setup

Instrument used for loading the vehicle.

Chassis dynamometer will have a gen/motor unit.

For loading it will be in generator mode and the

load is given in terms of opposing current.

Instrument used for measuring fuel

consumption.

The fuel flow rate was measured by timing the

consumption for 10 cc quantity of fuel from a

burette connected to the fuel flow line.


Fig 5.volumetric efficiency comparison

Figure 5 shows the volumetric comparison before

and after modification. From this it is inferred that

the volumetric efficiency has increased to the

maximum of 3.5%.

Fig 6.comparision of TFC before and after

modification for a load of 100N

Fig 7.comparision of TFC before and after

modification for a load of 150 N

From fig 6 it is inferred that the TFC for a load of

100 N has reduced because of the increase in

volumetric efficiency. Increase in volumetric

efficiency increases the cylinder pressure thus

compensating the fuel.

From fig 7 it is inferred that the TFC for a load of

150 N has reduced because of the increase in

volumetric efficiency. Increase in volumetric

efficiency increases the cylinder pressure thus

compensating the fuel.

From fig 8 shows power delivered at wheels for a

load of 100 N and it is inferred that increase in

volumetric efficiency improves the engine torque

which in turn increases the power delivered.

0

1

2

3

4

0 50 100

TFC

kg/

hr

Vehicle speed in kmph

TFC for 100N loadTFC before modification

TFC after modification

40455055606570

0 50 100

v eff

speed kmph

VeffV eff before modification

V eff after modification

0

1

2

3

0 50 100

TFC

kg/

hr

Vehicle speed in kmph

TFC for 150 N loadbefore modification

after modification

59



Fig 8.comparison of power at wheels before and

after modification for a load of 100 N

Fig 9.comparison of power at wheels before and

after modification for a load of 150 N

From fig 9 shows power delivered at wheels for a

load of 150 N and it is inferred that increase in

volumetric efficiency improves the engine torque

which in turn increases the power delivered.

CONCLUSION

From the results of the vehicle performance test

before and after modification, it is inferred that

using the frontal pressure of the vehicle which is

generated on the vehicle movement the volumetric

efficiency is increased and also has an effect on

vehicle performance.

This concept of improving the volumetric

efficiency can be used in vehicles with some

improvements, so that the turbo or super charger

can be eliminated thereby reducing the vehicle cost.

REFERENCES

1. Domkundwar Anand V ―Internal Combustion

Engines‖ (SI UNITS), IST ED 2007, Dhanpet

Rai & co

2. R.B.Gupta ―Automobile Engineering‖, 2005.

Satya prakashan, New Delhi

3. V. Ganesan, ―Internal combustion engines‖,

2003, Tata McGraw-Hill

4. John B. Heywood, ―Internal Combustion

5. Engine Fundamentals‖, 2000, McGraw-Hill

6. Richard stone ―Introduction to Internal

Combustion Engine‖ Third Edition, 2006,

Korean Society of Mechanical Engineers

7. V.Sumantran, Gino Sovran,‖ Vehicle

aerodynamics‖ 2003, Society of Automotive

Engineers

8. R.K. Bansal ―A textbook of fluid mechanics

and hydraulic machines‖ (in S.I. units), 2005

by Firewall Media

9. S. M. Yahya ―Fundamentals of Compressible

Flow‖, 2003, New Age International

10. www.sae.org

CONTACT

[email protected]

0

0.5

1

1.5

2

0 50 100

po

we

r at

wh

ee

ls in

kw

vehicle speed in kmph

power at wheels for 100N load

before modification

after modification

0

1

2

3

0 50 100

po

we

r at

wh

ee

ls in

kw

vehicle speed in kmph

power at wheels for 150N load

before modification

after modification

60



N-2010-E01-18

Experimental Investigation on the performance and

emission of Di Diesel Engine using Eucalyptus and Diesel

in dual fuel mode

*V. Pandiyarajan, Mr. V. Nadana Kumar, Dr. G. Sankara Narayanan

Keywords: Biodiesel; eucalyptus single cylinder diesel engine Exhaust emission.

ABSTRACT

An experimental investigation was carried out with

biomass-derived eucalyptus oil with diesel as a fuel

for diesel engine. The properties of eucalyptus oil

like viscosity, volatility, calorific value, latent heat

of vaporization, boiling point are comparable with

diesel fuel; hence it has been tried in diesel engine.

However, the low cetane number of eucalyptus oil

had prevented the 100% replacement of diesel with

eucalyptus oil in diesel engines. In the present

work, it was found that the engine was able to run

with eucalyptus oil and diesel blend up to the ratio

of 60:40 on volume basis. But due to erratic

performance and undesirable noise, the ratio was

restricted to 50:50. This has been done to quantify

the amount of diesel fuel replacement and to

maintain the minimum requirement of cetane

number. Smoke, UBHC and CO emissions were

decreased with slight increase in NOx for

eucalyptus-diesel blends. There was an increase in

brake thermal efficiency for all the blends at higher

loads without much variation at lower loads.

Combustion characteristics like heat release rate

and ignition delay were calculated and compared

with standard diesel. Combustion characteristics of

Eu 20 blend was in close agreement with standard

diesel. Based on the performance and combustion

analysis, Eu 20 blend has been proposed for diesel

engine operations.

INTRODUCTION

With over six billion people and 600 million cars in

the world today, the global energy requirement is

skyrocketing. From the increased pressure from

international initiatives such as the Kyoto

Agreement to reduce carbon emissions and the

lobbying activities of environmental pressure

groups, it is clear that governments have a tough

challenge on their hands.

Bio-fuels, namely, vegetable oils can be used as

fuels for diesel engines. Vegetable oils can be

directly used in diesel engines as they have a high

cetane number and calorific value, which are very

similar to those of diesel. However, the brake

thermal efficiency of vegetable oils is inferior to

that of diesel. This leads to problems of high

smoke, HC and CO emissions. This is because of

the high viscosity and low volatility of vegetable

oils, which lead to difficulty in atomizing the fuel

and mixing it with air. Further, gum formation and

piston sticking under long-term use due to the

presence of oxygen in their molecules and the

reactivity of the unsaturated HC chains, present

problems in the use of vegetable oils.

These problems were overcome by chemically

altering the vegetable oil (transesterification) and

blending it with diesel and Transesterification of

vegetable oils results in better performance and

reduced emissions. This process needs either

ethanol or methanol. A specified amount of

methanol is mixed and allowed to react with the

vegetable oil in the presence of a catalyst like KOH

or NaOH at a temperature of 70 °C.

Transesterification of vegetable oil provides a

significant reduction in viscosity, thereby

enhancing its physical properties. The cetane

number is also improved. It has been reported that

the methyl ester of vegetable oils offers lower

61



smoke levels and higher thermal efficiencies than

neat vegetable oils.

Further, it has been reported that the thermal

efficiency of the engine increases with an increase

in the methanol fraction in diesel due to an

increased fraction in the premixed combustion

phase with marked reductions in CO and HC

emissions. A marginal increase in NOx emission

and a reduction in CO, HC and smoke, due to the

presence of oxygen in neat bio-diesel and bio-

diesel–diesel blends were recorded and reported.

The behaviour of the bio-diesel prepared from

modified feed stocks was studied and it was

reported that the engine performance and

combustion process of all the blends were similar

to those of diesel fuel with marginally higher fuel

consumption, a shorter ignition delay, and a lower

premixed burning rate. The effects of cetane

numbers and fuel injection pressures on a diesel

engine emission and on its performance were

reported. The results showed that NOx, and CO

emissions reduced by about 15% and 5%,

respectively, when the fuel CN was increased for

standard injection pressure, but the smoke value

increased dramatically when the injection pressure

was reduced to 100 bars.

EXPERIMENTAL INVESTIGATION

Untreated eucalyptus oil is mixed with a mixture of

anhydrous methanol and a catalyst (NaOH) in

proper proportion. The mixture is maintained at a

temperature little below 650C and continuously

stirred the mixture for around three hours. After

completion of stirring, the mixture is allowed to

settle down for 24 hours. The layer of glycerol

settled at the bottom is carefully taken out and the

upper layer is the ester of eucalyptus oil which is

tapped separately.

Properties:

Bio Diesel Preparation:

Figure 1 Bio Diesel Manufacturing

Experimental setup and test procedure:

To study engine performance and emission, the

experiments were done in Kirloskar make single

cylinder, direct injected compression ignition water

cooled diesel engine (Engine model – AVI).

Figure 2 Experimental set up

62



The power output of the engine is 3.3 KW @

1500rpm having compression ratio 16.5:1. The

emission as well as engine performance was

studied at different engine loads (25, 50, 75 and

100% of the load corresponding to load at

maximum power.

Figure 3, 4, 5 Engine speed with specific fuel

consumption

Brake Thermal Efficiency:

The Brake Thermal efficiency varies with load, for

different fuel blends. Brake thermal efficiency is

increased due reduced heat loss with increased in

load. The maximum efficiency obtained in this

experiment was 33.74% (B25) and 33.54% (B20).

But considering the viscosity B20 is the better

option and this value is comparable with the

maximum brake thermal efficiency for diesel

(34.45%). From fig: 6, it is found that brake

thermal efficiency for biodiesel in comparison to

diesel engine is a better option for part load on

which most engine runs. The variation of BSFC at

different load and with the different brake thermal

efficiency is shown in figure 5 and fig: 6. for all

cases BSFC decreases with increase in load.

The reverse trend in the BSFC may be due to

increase in biodiesel percentage ensuring lower

calorific value of fuel. Another reason for the

change in BSFC in biodiesel in comparison to

petrodiesel may be due to a change in the

combustion timing caused by the biodiesel‘s higher

cetane number as well as injection timing. At

quarter load BSFC reduces a minimum of 3.2%

(B10) and a maximum of 5.85% (B5). For full load,

BSFC increases at a minimum of 5.66% (B5) and a

maximum of 20.82% (B10).

Figure 6 Thermal efficiency

ENGI NE EMISSION

CO emission: The variation of CO produced with

diesel and Diesel blends are presented in fig: 10.

For B20 blend the maximum and minimum CO

produced is 0.42gm/Kw-hr and 0.05 gm/Kw-hr,

which is much less than that specified in EURO –

IV Norms (max 1.5gm/Kw-hr). It is an indication

of the complete combustion of biodiesel being an

oxygenated fuel.

63



Figure 7 Setup

NOX Emission:

The variation of NOX at different engine load is

presented in fig: 8. the reason for the increase in

NOX is not clear. The cetane numbers of the

biodiesel are generally higher than that of diesel

fuel associated with lower NOX emission. The

injection timing advancement associated with these

effects could be partially responsible for the

increase in NOX emissions. For B20 blend the

maximum and minimum NOX produced is

0.04gm/Kw-hr and 0.002 gm/Kw-hr, which is

much less than, mentioned in EURO – IV Norms

(max 3.5gm/Kw-hr).

CO2 Emission:

The variation of CO2 produced at different engine

load is presented in fig: 9. For B20 blend the

percentage increase for minimum and maximum

load is 3.8 and 3.75. This increase in percentage

may be due to complete combustion of the fuel.

Figure 8 NOx emission

Figure 9 CO2 emissions

Hydrocarbons:

The variations of un-burnt hydrocarbon at different

engine load for different diesel blends are shown in

fig: 10. the shorter ignition delay associated with

biodiesel higher cetane number could also reduce

the over mixed fuel which is the primary source of

un-burnt hydrocarbons. For B20 the maximum and

minimum HC produced is 0.02gm/Kw-hr and 0.004

gm/Kw-hr, which is around same as that is

64



mentioned in EURO – IV Norms (max 0.02

gm/Kw-hr).

Figure 10 Hydrocarbons

Exhaust temperature:

The exhaust temperature variation profile for

different diesel blends are shown in fig: 11. for

diesel engine the minimum and maximum exhaust

temperature is 1200C and 260

0C, whereas for B20,

it is found to be 810C and 216

0C.

Figure 11 Exhaust temperature

CONCLUSION

1. Based on engine emission studies i.e. CO, NOX

and hydrocarbon, we can say that all the parameters

are within maximum limits that conclude safer use

as an alternate fuel.

2. Eucalyptus oil can be one of the hopeful

alternatives for diesel engine.

3. The blending ratio of eucalyptus to gas would be

less than E40/G60 (eucalyptus 40 to gas oil 60

volume ratio), the knocking tendency was

recognized in case of higher blended eucalyptus oil.

4. The value of e2o/g80 would be a desirable blending

ratio for diesel engine.

5. Ignition improving agent will except for the engine

startability, in case of using eucalyptus oil for

diesel.

REFERENCES

1. Agarwal AK. Vegetable oils verses diesel fuel:

development and use of biodiesel in a compression

ignition engine. TIDE 1998; 8(3):191–204.

2. Choi, C. Y., Bower, G. R. and Reitz, R. D., 1997.

Effects of Biodiesel Blended Fuels and Multiple

Injections on D.I. Diesel Engine, SAE Paper No.

970218: 388-407.

3. Harrington KJ. Chemical and physical properties of

vegetable oil esters and their effect on diesel fuel

performance. Biomass 1986; 9:1–17.

4. Masjuki H, Salit. Biofuel as diesel fuel alternative:

an overview. J. Energy Heat Mass Transfer 1993;

15:293–304.

5. Peterson, C. L., Thomp-son, J. C., Taberski, J. S.,

Reece, D. L. and Fleischman, Q., 1999. Long-range

on road test with twenty percent rapeseed biodiesel.

Applied Engineering in Agriculture, ASAE Vol.

15(2): 91-101.

6. Piyaporn K, Narumon J, Kanit K. Survey of seed

oils for use as diesel fuels. J. Am. Oil Chem.Soc.

1996; 71(4):471–7.

7. Ryan TW, Dodge LG, Callahan TJ. The effects of

vegetable oil properties on injection and

combustion in two different diesel engines. J. Am.

Oil Chem. Soc. 1984; 61(10):1610–9.

8. Sinha S, Misra NC. Diesel fuel alternative from

vegetable oils. Chem. Engng World 1997;

32(10):77–80.

9. Srivastava, A. and Prasad, R., 2000. Triglycerides-

based diesel fuels. Renewable and Sustainable

Energy Reviews, Vol. 4:111-133.

10. Ziejewski M, Kaufman KR. Laboratory endurance

test of a sunflower oil blend in a diesel engine.J.

Am. Oil Chem. Soc. 1983;60(8):1567–73.

CONTACT * Department of Mechanical Engineering,

AdhiParasakthi Engineering College,

Melmaruvathur – 603319

[email protected]

65

mailto:[email protected]



N-2010-S04-21

Suitability of Multi - Core for Embedded Automotive

System

Santosh Kumar Jena, Dr. M. B Srinivas

IIT Hyderabad

Key words: Embedded system, automotive safety, processor

ABSTRACT

This paper describes the suitability of multi-core

controller over single core in automotive safety

critical applications. As vehicles become more and

more complex and an embedded network

interconnection of ECUs (Electronic Control Unit)

integrates more and more features, safety

standardization is becoming increasingly important

among automakers and OEMs (Original Equipment

Manufacturer). It is also necessary to improve the

processing power to meet all requirements of time

critical functionalities. Multi-core processor

hardware is seen as a solution to the problem of

increasing ECU processing power with the support

of software. Here, ABS (Anti-Lock Braking

System) is taken as an example and the timing

issues in hard braking system are described. In this

work, it is shown how use of multi-core processor

helps to overcome it.

INTRODUCTION

According to Mark Fitzgerald [1], embedded

control will continue to be a primary driver of high-

end automotive microcontroller performance and

functional development. However, going forward,

multi-sensor advanced driver assistance system

applications will increasingly emerge with high

performance processor drivers. As a result, multi-

core designs will increase for greater computational

performance.

In this paper, the authors explore the application of

multicore computing for automotive embedded

applications and the performance of Anti-Lock

Braking System (ABS) has been studied with the

help of TMS570 which is a dual core controller

from Texas Instruments and compared with that of

TMS470 which is single core controller from the

same manufacturer. A software architecture using

MPI (Message Passing Interface) is described in

detail and applied to quantify the performance.

Automotive Electronic & Abs

Auto Mobile Electronic controllers Units (ECU)

are divided into 3 categories.

1. Power Train Controllers.

2. Chassis Controllers.

3. Body Controllers.

The Braking system is with Chassis Controller.

Brake Controllers include ABS, which is the first

slip control technology. This allows the brake

pressures at each wheel to be modulated to prevent

the wheels from locking, thereby maintaining

steerability, and reducing stopping time and

distance.

When slam on the brakes, the sensors sense that

wheels are slowing down. If one of the wheels is

about to stop rolling, the ECU will separately work

the brakes at each front wheel and both rear

wheels. The antilock system can change the brake

pressure faster than any driver could. The ECU is

programmed to make the most of available tire &

road condition. As long as the brake is applied, the

ECU keeps receiving updates on wheel speed and

controls braking pressure accordingly.

The advantages of ABS are it

1. reduce the stopping distance

2. improve stability

3. improve steerability during braking

66



ANALYSIS AND PHYSICS FOR ABS

The wheel rotates with an initial angular speed that

corresponds to the vehicle speed before the brakes

are applied. Separate integrators are used to

compute wheel angular speed and vehicle speed.

Slip, which is determined by Equation-1.[5] is

calculated for two different speeds.

Vehicle speed can be expressed as an angular

velocity.

ωv = V / R (Equals the wheel angular speed is

there is no slip) ------ (1)

ωv = Vv / Rr

Slip = 1 – ωw/ ωv

ωv = Wheel speed divided by wheel radius

Vv = Vehicle linear velocity

Rr = Wheel radius ωw = Wheel angular Velocity

From these expressions, we see that slip is zero

when wheel speed and vehicle speed are equal, and

slip equals one when the wheel is locked. A

desirable slip value is 0.2, which means that the

number of wheel revolutions equals 0.8 times the

number of revolutions under non-braking

conditions with the same vehicle velocity. This

maximizes the adhesion between the tire and road

and minimizes the stopping distance with the

available friction.

ADDRESSING REAL-TIME PROBLEM

In the case of hard braking or sudden fall of vehicle

speed due to accident or collision, the following

cases may happen:

1 System has to identify the reason of

sudden deceleration as accident, collision,

or due to hard braking.

2 This information is needed by braking

system, driver information display, airbag,

engine controller.

3 If due to hard braking, it also needs

information about steering angle, wheel

speed, brake pressure, engine status,

lateral speed, yaw rate and longitudinal

speed.

4 Airbag controller needs collision

information, seat belt, passenger

occupancy etc.

5 Driver info display may not be mandatory

in this case but for future investigation the

information should be stored.

6 Engine speed has to be reduced.

But, with the overloading of task and increasing

number of sub-systems, the load on the processor

has reached an optimum level. This, often, leads to

increase in the waiting time of a task. With this, it

may not always be possible to meet the real time

requirements. To run the system, it is mandatory

that the entire task should be completed within the

schedule time, instead of waiting for other tasks.

To increase the safety, the system is adding more

sensors to understand environmental situations, and

more features to support and to reduce the number

of ECUs in the vehicle, which makes the tasks run

faster.

MULTICORE COMPUTING & MPI

To address the above real time problem, the

architecture was re-designed with Texas‘s

TMS570, a dual core single chip microcontroller

[2]. In TMS570 the two core is Cortex M3 and

Cortex R4. Each is treated as an individual

processor. Each processor has the capability to

send an interrupt the other processor for message

notification via: MPI.

MPI is a standard for communication between

processes on a distributed memory system, and is

implemented on shared memory system. It

supports point to point as well as collective sending

and receiving of messages.

The MPI is generally used in a desktop application

with interconnection of CPUs in a network,

considering each as an individual processing unit.

But it is a challenge to implement MPI in an

embedded application which supports multicore

processor.

Both cores maintain an array of ongoing ‗send and

receive‘ requests messages Via IMM (Inter-

processor Messaging Module). Each communicator

67



receives a notification that the receive request fifo

contains a new request. This activates the MPI

Task. The communicator task pulls the data out of

the fifo and inserts that request into its array of

ongoing requests. It stores that information in a

field which is reserved for a predefined request. In

case the previous request has not been completed,

the corresponding send or receive routine shall

return an error.

In general, as given in the example given below,

there are a couple of tasks waiting for any transfer

being completed. IMM is used to send out

notification messages to trigger an MPI scheduler,

which is responsible to unlock waiting task upon

completion of any transfer.

Figure1. IMM Interface.

The communication always works in a way that

any posting is completed when both the sender and

receiver have posted their requests and the data

transfer is finished. The communicator task

compares the send and a receive requests. In case

of a match the communicator, who has posted the

send request, copies the data from the source buffer

to the target buffer and tags the send and receive

request as completed. Additionally, it posts the

receive request via IMM back to the receiving

communicator with the complete flag being set.

Due to the static design, all requests will be stored

at the same location. In order to distinguish

between a newer posted request and a previous

request a sequence number shall be maintained.

That means that rank, communicator, tag and

sequence number must match in order to process

the request. The default sequence number is 0; the

default state for the completion state is 1.

Once the receiving communicator receives this

request he will tag it‘s receive request to be

completed as well. On both sides the corresponding

request buffer is available again and can be used

for the next requests. Both Cores need to share

their data via shared memory. The M3 can store its

data into its local memory, which is the shared

memory, the send or receive address may be the

same address when processing the data after

transfer.

The R4 is capable to send data directly to the M3s

local ram space. If R4 is the receiver M3 can put its

data into shared memory only since M3 has no

direct access to the R4s local memory space. This

may be application specific.

If receive and send communications are the same, a

task sends data of another task (or its own task).

For this purpose it is not needed to send out any

IMM message. This can be handled using regular

OSEK service routines (SetEvent). The MPI

handler will take this into account and fires the

alarm on the same core (communicator).

Since the TMS570 has no means for mutual

exclusive access across both cores each core has to

maintain a local copy of all requests. This copy

holds the identification parameters needed to

access the detailed information in the shared ROM

table.

Independent from a ‗receive or send‘ request, a

posted request has the following format:

Width Value

11 RESVD

2 COMM

4 RANK

8 TAG

3 SEQ

3 ERR

CMPLT

RESRVD

Reserved for future purposes

COMM:

Communicator [0..3] : depending on a receive

or send request this field specifies the source

or target communicator id

RANK

Rank, [0.number of tasks within the

communicator].

68



Depending on a receive or send request this

field specifies the source or target rank in a

given

communicator

TAG

Tag [application specific value]. This field

may contain some user defined value to

distinguish between

several requests going to the same node of a

communicator

SEQ

Sequence Number: implements a module

counter to distinguish a newer request from a

previous one which hasn‘t completed so far.

ERR

Error state, provides some error codes which

are passed across the nodes. Could also be

used to cancel a request, which has already

been posted.

CMPLT

Completed flag. Set after a request has

completed.

The 32 bit word above will be send via IMM to

either core whenever a new request needs to be

posted.

An IMM message will not be sent when receive

and send communicator are the same.

R4 / M3 Memory Mapping

A generic table needs to be allocated in shared

ROM, so that both cores can access the static

configuration table without the need to duplicate it

for each core. The table is declared and defined in

R4; the M3 needs to know the structure and the

location in its ROM area but doesn‘t allocate

memory for this.

Each entry in ROM has a pointer in shared memory

to a 3 entry ram pointer table. This table stores the

addresses of the corresponding send / receive and

buffered send buffer of each posted request. The

addresses will be stored during runtime from each

posted request, since the addresses of any send or

receive buffer is not known on the other core

during link time. Note that addresses in shared

memory (either ROM or RAM) are accessible via

TLB and must have an offset for R4 to access the

memory. If the R4 detects that the receive buffer is

in shared memory all access needs to go through

the TLB via an offset. The send buffer itself can be

located in R4 or M3 (shared RAM).The receive

buffer address for R4 must be in shared RAM, the

receive buffer addresses for M3 can be located in

the entire M3 RAM space (which is shared RAM

and all accessible by R4).In case that M3 posts a

receive request to R4 the receive buffer address

shall have the appropriate offset which is stored in

the receive buffer table. In case of a send request

no TLB offset is needed, since all data goes into

M3 internal memory. In case that R4 posts a send

or receive request the address has already a TLB

offset, since the source/destination address is

already given by the linker.

Given below is a scenario where R4 posts a send

request to M3 and M3 posts a receive request to

R4.Both cores need to read the configuration table

to figure out if the received requests are valid (e.g.

if they match with the static design) and obtain the

addresses where to store the send / receive buffer

addresses. During send operation R4 needs to add

the appropriate offset when accessing M3s receive

buffer.

Post send request

#3

Post recv request

#3Confi. Request #3

Get config. data

Store recv buff addr

Get config. data

Store send buff addr

Confi. Request #2

Confi. Request #1

Confi. Request #n

Confi. Request #...

buffer addr req #1

buffer addr req #2

Send

recv

bsend

Buffer addr req #3

buffer addr req #...

buffer addr req #n

R4 M3

Points to... Points to..

rom

Shared ram

MPI_TransferData

Receive address +=

TLB offset

Send bufferReceive

buffer

Figure2. M3 / R4 Send Receive mesg

69



Implementation MPI Functions

MPI_Send

This routine posts a blocking send of data from

buffer to a node with rank rank in communicator

comm.

Buffer is a ram location which may contain any

data (U8..U32, arrays, structs, …). Rank is the

(MPI-) naming for a node, which is the task id in

this case. Communicator is the node (more

precisely: the core) to which data to be send to.

MPI_Send first locks the MPI interface to prevent

from data corruption during access to the data

structures.

If a matching RECV request has already been

posted (e.g. rank, tag, communicator and sequence

number match) the SEND request will be stored in

the request array and a SEND request is posted via

IMM. MPI_Send will then refer to the const data

structures to obtain destination and source buffer

address as well as the element size and the type of

each element. Then it copies the data from the

source buffer into the destination buffer if the

addresses are different. Once this is done the

request complete flags will be set and a RECV

complete notification is send out via IMM to force

a rescheduling on the communicator who has

posted the RECV request. If no matching RECV

has been posted so far, MPI_Send checks if the

previous RECV request has been completed.

MPI_Send unlocks the MPI interface and returns

an error if the previous operation is still ongoing

and has not completed so far. It‘s up to the

application to retry.

In case that the previous request has completed

MPI_Send clears the complete flag, stores the

SEND request with tag, rank, communicator and

updated sequence number in to the request buffer

and posts a SEND request via IMM.

The MPI interface needs to be unlocked since other

messages may need an update due to new IMM

notifications (via interrupt). MPI_Send waits for

the reception of a matching RECV request through

WaitEvent(). Once SEND and RECV requests

match MPI_Send locks the MPI interface, reads the

source and destination addresses from the

corresponding ROM table and copies all elements

into the receive buffer. Upon completion it sets the

request complete flags and posts a copy of the

current RECV request via IMM back to the

communicator who has posted the RECV request.

This message has the complete flag being set. The

receiving communicator will take that information

to finally unlock the task which has posted the

RECV request. Before returning to the caller

MPI_Send unlocks the MPI interface again. The

function returns with no error.

In case that the sender and receiver are located on

the same communicator no IMM action takes

place. The message can be copied directly to the

receiving task (which in theory can be the same

task). Note that this needs to be a buffered

MPI_BSend and a MPI_Recv call to prevent from

a deadlock. In case that the receiver is a different

task on the same communicator MPI_Send and

MPI_Recv can be used more safely.

MPI_Recv

This routine posts a blocking receive of data from

buffer to a node with rank rank in communicator

comm.

Buffer is a ram location where the receive data will

be stored and may contain any data (U8..U32,

arrays, structs, …). Rank is the (MPI-) naming for

a node, which is the task id in this case.

Communicator is the node (more precisely: the

core) from which data are to be received from.

MPI_Recv first locks the MPI interface to prevent

from data corruption during access to the data

structures.

If a matching SEND request has already been

posted (e.g. rank, tag, communicator and sequence

number match) the RECV request will be stored in

the request buffer and the RECV request is posted

via IMM. The MPI interface needs to be unlocked

now since other messages may need an update due

to new IMM notifications (via interrupt).

MPI_Recv now sticks in an event routine and waits

for the reception of a matching RECV complete

request. It then returns with MPI_SUCCESS.

If no matching SEND request has been posted so

far MPI_Recv checks if the previous RECV request

has been completed. MPI_Recv unlocks the MPI

interface and returns an error if the previous

70



operation is still ongoing and has not completed so

far. It‘s up to the application to retry.

In case that any previous request has completed

MPI_Recv clears the complete flag, stores the

RECV request with tag, rank, communicator and

updated sequence number in to the request buffer

and posts a RECV request via IMM.

The MPI interface needs to be unlocked now since

other messages may need an update due to new

IMM notifications (via interrupt). MPI_Recv sticks

in an event routine and waits for the reception of a

matching RECV complete request. It then returns

with MPI_SUCCESS.

MPI_TransferData

MPI_TransferData is the push routine which sends

data to the other node or communicator. The

transfer routine ensures that the correct TLB offset

is applied to the receive buffer address in case that

the receive buffer is located in M3 address space

and R4 sends data.

Truth table for TLB offset correction:

send \ recv R4 mem M3 mem

R4 mem No TLB offset TLB offset

M3 mem No TLB offset No TLB offset

MPI_TransferData

nono

yes Target == M3

Communicator == R4 yes

Receive buffer

address += TLB

offset

copy data to

receive buffer

Figure3. MPI_Transfer Data

A number of elements and data types are passed as

arguments to this service. This will be used to

optimize the transfer in terms of word length.

MPI_Scheduler

MPI_Scheduler runs on both cores as a basic task

which supports multiple-activation and is

responsible for activating/unlocking Tasks due to

blocking receive / send MPI calls. This task is

activated on IMM receive events.

Once activated it reads the message buffer from

IMM which contains communicator id, rank id, tag

and some states to determine which receive or send

MPI call is blocking. Having done this, MPI Task

terminates.

MPI_Task

Valid messageno

Process received data

yes

Inform

Faultmanager

IMM receive event

TerminateTask

Set MPI Event

Figure4. MPI_Scheduler Algorithm

SOFTWARE DESIGN

The software interacts with

1 The Algorithms & non performance

software

2 The Brake system model.

HW model sends the valve commands

and the pump-motor commands, to

the Braking system model.

The Braking system model

communicates the MCP to the

Hardware model.

3 The Vehicle model.

The HW model receives from the

vehicle model.

4 The CAN data such as Steering angle,

YAW rate, longitudinal and lateral

71



acceleration, Engine Torque, Engine

RPM, Throttle position

5 Wheel speeds

The HW model sends to the vehicle

model

6 Engine Torque Request through the CAN

link

Braking System

ModelVehicle Model

Driver Model

Surface Model

Algorithm S/W

Wheel Speeds

CAN DATA

Steering Angle sensor

Yaw rate sensor

Lateral acceleration

Longitudinal acceleration

Vacum pressure sensor pri

Vacum pressure sensor sec

Throttle position

Valve Commands

Pressure Motor

Master Cylinder

Pressure

S/W Flow diagram

SCHEDULING AND TASK ORGANIZATION

The braking application software is designed as a

set of atomic schedule items, scheduled

synchronously. Atomic in the sense that the

execution of one schedule item cannot be

interrupted and the execution of another atomic

schedule item cannot be scheduled. Hence the new

schedule is designed for such an application to

function correctly in a preemptive environment.

Such a solution comes with an execution hit. As the

braking software migrates to the preemptive

scheduling environment, care is to be taken to

avoid the scheduling overheads where ever we

could.

With the existing software, the atomic schedule

items are defined as Update…() calls. The

execution of all the Update…() calls can in

interleaved in any order, as long the periodicity of

the Update…() calls are met. The Update…() calls

were traditionally designed to not exceed 100ųs of

execution time.

In the OSEK (Open Systems and their

Interfaces for the Electronics in Motor

Vehicles) we are using three tasks:

1 HighTask gets called for example

periodically every 1ms. This task has the

highest privilege level and can‘t get

interrupted by any other task.

2 MidTask gets called for example

periodically every 5 ms and 7ms. This task

has the second highest privilege level and

can only get interrupted by the HighTask.

3 LowTask gets called for example

periodically every 10 ms and 14ms. This

task has the lowest privilege level and can

only get interrupted by the HighTask and

MidTask.

Scheduling Graph

The diagram below shows the execution of the 3

major tasks in the application. Namely the one

millisecond task (High priority task), the task

running at SCHEDULE_LOOP_TIME (= 5ms and

7ms) and 2 * SCHEDULE_LOOP_TIME.

Preemptions of the tasks are by interrupts.

1 millsecond

SCHEDULE_LOOP_TIME (5ms &

7ms)

Twice SCHEDULE_LOOP_TIME

Task Period

Time in milliseconds

t t+1ms t+2ms t+3ms t+4ms t+5ms t+6ms t+7ms t+8ms t+9ms t+10ms t+11ms

PASSIVE or

WAITING or

SUSPENDED

Ready Running

Legends representing task states

Figure5. Scheduling of Operating System

The application software cannot be preempted

when in the middle of an Update..(). Hence all the

Update…() calls with this constraint are scheduled

by disabling preemption before invoking the

Update..() call & re-enabling the preemption after

the execution of the Update…() call. Thus a given

task would have several Update…() calls and each

of these calls could be non-preemptable, however

the task can be preempted when in-between

Update…() calls. This is made possible by getting

the RES_SCHEDULER resource before the

Update…() call and releasing the resource after the

Update…() call. This mechanism of keeping the

72



legacy Update…() calls atomic schedulable

entities, comes with the overhead of the resource

operation. The whole process is made possible via

the implementation of a Sequencer and Sequence.

Sequence is composed of a list of Update…() calls

and their properties. Sequencer is a generic

function for the execution of a task merely

composed of a sequence. The properties of each

Update…() call in a sequence.

PCP (Priority Ceiling Protocol) is used to make

special sections or function calls non preemptive to

avoid data loss. Therefore we defined two

resources with different priority increase levels.

resource_high: this resource increases the

Priority of the current task to the highest priority.

When using this resource inside the task to block a

function, the function can‘t get interrupted by any

other task.

resource_middle: this resource increases the

Priority of the current task to the middle priority. If

you use this resource inside LowTask, LowTask

can‘t get interrupted by MidTask but HighTask can

still interrupt.

Also along with some exception used smoother

flow of data and running of the system

1. Function calls that should be executed

periodically with a minimum of jitter, for

example hardware IO functionalities, must

be called in the beginning of the task.

1 Local variables used in a subsystem shall

not be used by other subsystems, except

for data logging.

2 Fault Module requires locking and

protection. This data resource is being

accessed by multiple tasks.

3 Data transfer between tasks at different

rates should be done by polling. A task

shall not push data into another task of a

different rate.

4 Input and output buffers should be used

for data synchronization in the slower rate

tasks between tasks of a different rate.

Data used between subsystems run in the

same task does not need to be buffered, to

minimize RAM usage. The Input and

Output functions shall be protected against

preemptions to avoid data loss via the

operating system.

RESULTS

With the use of faster CPU processing and

portioning the tasks among the cores, the ABS

performance result has enhanced. Below table

shows the comparison between the result received

from TMS470 and TMS570 based ABS system [3].

Table1. Results

Vehicle

Speed

(km/hr)

TMS470

stopping time

( Sec )

Slip = 0.2

TMS570

stopping time

( Sec )

Slip = 0.2

TMS570

stopping

time In

Real Road

60 1.819 0.645

80 1.43 0.619

100 1.85 1.31 2.656

The results are the simulated value and it may vary

in real time, as it depend the friction of road, yaw

rate and the throttle poison of the vehicle.

Analysis / Graphs

Fig6. The results for TMS470 @ vehicle speed 80

km/sec and stopping time aprox 1.43 sec

73



Fig7. Vehicle speed 100 km/sec and topping time

aprox 1.85 sec

Fig8. The results for TMS570 @ vehicle speed 80

km/sec and topping time aprox 0.619 sec


aprox 1.31 sec


aprox. 2.656 sec in the actual road

CONCLUSION

The use of multi-core processor will increase in

automotive applications because of the growing

need for increased performance with lower power

consumption and adding more safety

functionalities.

The need to reduce the number of ECUs in

vehicles, and the interconnections between them,

will lead to the emergence of new, more centralized

architectures that are more efficient and reliable,

less complex, and more cost-effective.

This migration will require changes in software

architectures and mapping (e.g., OS, libraries,

drivers), and will need new tools to support the

development, reuse of legacy code, validation and

qualification. The long-term result, however, will

be expanded vehicle performance with vastly

improved safety and unmatched comfort.

Multi-core is innovative approach to safety.

Meeting the performance demands of future high-

end applications will demand more multicore

systems, and this will require new verification

methodologies for the development of new

hardware and software.

REFERENCES

[1] http://www.embedded.com/design/multicore

[2] http://www.ti.com/TMS570

[3] Test results from CMC Limited, Automotive

74

http://www.edn.com/hot-topic/48833/multicore.html



N-2010-E08-22

GANESH for SI Engine Simulations – GUI Approach 1 Vijayashree,

2 Tamilporai P,

3 Ganesan V

Keywords: SI Engine, Simulation, GUI, Combustion analysis

ABSTRACT

This paper deals with the development of a

graphical user interface (GUI) to simulate

progressive combustion process in spark (SI)

ignition engines using thermodynamic equations.

GUI has been developed using Matlab to predict

the performance of four-stroke SI engines, and is

named as GANESH. It is the acronym that stands

for Graphical and Numerical Engine Software Hub.

Using the developed software, the trends of the

performance, as a function of some important

engine variables can be simulated. A limited

amount of validation has been done. In addition,

parametric studies have been carried out to analyze

the engine performance. The results obtained by

solving appropriate process governing equations

are presented in the form of graphs.

Most difficult aspect of the engine simulation is to

predict the pressure-volume variation during the

combustion process and to compute the power

output. For progressive combustion simulation,

three simple thermodynamic models have been

used. The three models are: Uniform rate model,

which depicts a combustion chamber, which can

produce a uniform rate of reaction; Square law,

which predicts the continuous growth of the flame

front area with initial slow growth; Cosine law

model, an empirical law derived from the

experimental studies which depicts reasonably the

realistic progress of combustion

Computer simulation models developed here gives

results consistent with the experimental data and it

may be used as a tool for the performance

evaluation of a spark ignition engine. It can pave

way for viz. better understanding of the variables

involved and their effect on engine performance.

Thereby, it reduces considerably the time-

consuming tests by narrowing down the variables

involved for development. Thus, all processes

involved in a four-stroke SI engine are simulated.

The user friendly software developed can be used

with ease and will be particularly useful for getting

results which will reduce the development time.

INTRODUCTION

Computer simulation has become a powerful tool,

which is the most economical process as compared

to the experimental study and time minimization. A

proposed theory can be analyzed quickly using a

computer and the cost of setting up costly

experimental apparatus can be postponed until

optimization is achieved. The main concern to

develop a computer model for an engine is to

achieve the improvement of performance, viz.

increased efficiency and reduction in specific fuel

consumption and emissions. One of the major

thrust nowadays in the design and development of

the modern internal combustion engine is by means

of computer simulations of various processes Their

economic value is in the reduction in time and cost

for the development of new engines and their

technical value is in the identification of areas that

require specific attention as the design study

evolves.

However, it may be noted that the simulation is

only a step prior to experimentation and the results

obtained from simulation must be validated with

experimental results to establish the reliability.

Once validated, computer simulation can provide a

deep insight into the performance characteristics of

the system, which is true in case of internal

combustion engine studies.

In an internal combustion (IC) engine, the

processes involved are extremely complex. The

design of an engine relied heavily on previous

experience and know–how till the last decade. As a

75



result of this, the extensive testing of the prototype

was necessary for an engine development.

Selecting the best design from such testing is a task

of considerable magnitude. Because of these

difficulties, analysis by means of computer

simulation has become quite popular in recent

years. The theoretical models used in the case of

internal combustion engines can be classified into

two main groups, viz, thermodynamic models and

fluid dynamic models. Thermodynamic models are

mainly based on the first law of thermodynamics

and are used to analyze the performance

characteristics of engines. Pressure, temperature

and other required properties are evaluated with

respect to crank angle or in other words with

respect to time. The engine friction and heat

transfer are taken into account using empirical

equations obtained from the experiments. These

models are further classified into two groups

namely single – zone models and multi – zone

models. Multi–zone models are also called

computational fluid dynamics models. They are

based on the numerical calculation of mass,

momentum, energy and species conservation

equations in either one, two or three dimensions to

follow up the propagation of flame or combustion

front within the engine combustion chamber.

The present study is undertaken to develop a

computer code to simulate the SI engine processes

for various fuels. The model is developed in such a

way that it can be used for characterizing any

engine fuel, namely, petrol, diesel, methanol,

ethanol etc. The modelling results clearly indicate

that, with increase in compression ratio, peak

pressure, peak temperature and brake thermal

efficiency increases. The performance

characteristics of the engine follow the same trend

for all fuels. The predicted results are compared

with the experimental results of the engine fuelled

by gasoline. The simulation predicts the global

engine performance characteristics in closer

approximation to that of experimental results.

LITERATURE SURVEY

Simulation model and governing equations

developed by Ganesan [1, 2] has been adopted for

writing numerical procedure and the software.

Hayes [3], Craigen et.al [4], Bowen and Stavridou

[5] have applied formal methods to study the

numerous cases in industrial context. A number of

published works are available on engine modelling

(Lakshminarasimhan [6], Mathur et. al. [7],

Heywood [8]). The phenomenological approach to

model the physical processes relies on the

mathematical analysis of first law of

thermodynamics applied to an open system

composed of fuel-air-residual gas charge within

engine cylinders and manifold

Ball et al [9,10] have analyzed engine combustion

in detail to study its effect on engine performance

and emissions and have presented the reasons for

cycle to cycle variation of the engine performance.

Bazari et al [11] have proposed engine simulation

models with engineering building block approach.

Variables affecting the engine performance are

studied independent of each other.

Friction model of Bishop [12], which describes

empirically the magnitude of most important

factors determining the friction of an engine based

on experiments, is used to obtain the brake

parameters from indicated values. The friction

model takes into account mechanical friction in

crankcase, throttling losses, pumping loss and

friction due to piston movement.

Based on the literature survey it is seen that it is

worthwhile to simulate the actual engine

performance based on the thermodynamic approach

and to study the effects of various parameters on

engine performance, which will be useful as a

quick reference for students and practicing

engineers. It may be noted that simulation of

compression and expansion processes are

comparatively easier compared to intake and

exhaust processes since the former takes place

during closed period processes whereas the latter

takes place during the open period processes. The

most difficult process to simulate is the combustion

process. The important aspect of this paper is to

describe the combustion process by means of

thermodynamics to get reliable and quick results.

76



PROGRESSIVE COMBUSTION IN SI

ENGINE Combustion in S I engines can be simulated ideally

under constant volume. Constant volume adiabatic

flame temperature calculation can be incorporated

to evaluate the peak temperature of combustion.

However, it may be noted that in an actual engine

the combustion is progressive, meaning

combustion spreads over a period of crank angle

(time). The combustion of fuel and air mixture

inside the engine cylinders takes finite time

interval. The combustion duration and the rate of

heat release during combustion are most important

and they critically affect the peak pressure and

temperature there by the engine performance.

Ignition in SI engines starts before the piston

reaches top dead centre, and combustion is not

completed until the piston has moved beyond top

dead centre. Because of the progressive burning, it

is evident that the work done on the gases during

the compression process will be increased while the

work done by the gases during the expansion

process will be decreased. This will definitely

decrease the brake power output.

There are certain facts regarding the combustion

process in a spark ignition engine which are

important:

1. Combustion is initiated by a spark and after a

brief delay it continues by the movement of

flame front which propagates from the spark

plug at a finite rate.

2. Combustion is virtually complete when the

flame front passes through the entire charge

and reaches the opposite wall and gets

quenched.

3. The time required for combustion varies with

fuel composition, combustion chamber shape

and engine operating conditions.

4. Best power and efficiency are obtained when

crank angle at which combustion starts before

TDC is the same as that at which it ends after

TDC.

5. Due to knocking, the spark timing has to be

retarded.

6. Because of the moving flame front, with its

finite speed, it is evident that even if the piston

remained stationary during combustion (as

assumed with ideal combustion), different

parts of the charge would burn at different

times.

7. The above factors pertains to progressive

burning

It is based on the fact that the first charge burns

first and then gets compressed (that is, its pressure

rises) whereas the last charge first gets compressed

and then burns. This results in the temperature near

the spark plug (representing the first charge

burned) being higher than the temperature at the

farthest end (representing the last charge burned).

Theoretically this difference may be 600 K whereas

actually it may much lower (about 200 K).

However, theoretical analysis with and without

progressive burning, the mep and efficiency are the

same.

The main objective of this paper is to use GUI to

enable the user to input the values with ease and

also get the output in required format.

GRAPHICAL USER INTERFACE

The user interface is perhaps the most important

part of an application; it is certainly the most

visible. To users, the interface is the application;

they probably are not aware of the code that is

executing behind the scenes. No matter how much

time and effort one puts into writing and optimising

code, the usability of the application depends on

the interface.

A typical simulation engine framework can be

divided into a back end, which controls all

computations, and a front end which enables the

user to give input and get the required output.

Sometimes a third layer, called a visualization layer

if applicable, can be interfaced with these two. As

already stated, the responsibilities of a front end are

to control output and user input throughout a given

simulation. More specifically, it should accept any

input from the user or from files and use it to

initialise anything the back end needs to perform

computations. Then it should be able to access back

end information throughout a simulation and

77



perform appropriate output, be it to the console or

to files. In order to obey traditional front/back end

interfacing, which involves clear separation of the

two, this front end is designed not to perform any

computations whatsoever, and also not to force any

back end that it is interfaced with to do I/O.

Further, the front end has been made not to allow

changes to any back end that it is interfaced with to

affect its operation, outside of calling the

appropriate initialise and run methods that are

unique for each particular back end. The reverse is

also true, that changes to computational methods in

the back end should not affect the operation of a

front end. Thus numerous advantages in flexibility

are offered by this design, and they have been

proven in its reusability. In addition, there are

many other advantages that this design holds. User-

friendliness was an extremely important

consideration when creating this framework, and its

implementation was designed around file-based

input which is preferable to any user because it

means that one does not need to re-enter all input

parameters each time one changes. Finally, this

front end is designed to be easily modified.

Modifications can occur on existing programs often

by programmers or even users at times.

Since GANESH is developed under the GUI

environment, presentation is taken care to provide

output graphs by effective screen display. Screen

design includes:

Screen fields for user input and system

input

Display of calculated values

Selection of the type of screen fields

Presentation of required values

Plots

A typical user input GUI display screen is shown in

Fig.1.

Fig.1 GUI for Front end (User Input) – Engine Parameters

ABOUT GANESH

GANESH simulation code is an one-dimensional

thermodynamic general purpose engine

performance prediction software based on control

volume principle. Its underlying methodology

allows the simulation of the performance of spark-

ignition engines. Its main features include

a self-contained user friendly methodology

comprising pre-processing, analysis, post-

processing and reporting facilities

ability to implement flexible control

ability to access and control the simulation

through user programming

The block diagram, depicting the details of engine

simulation, is shown in (Fig.2).

GANESH is user friendly software. Typical

applications of which are:

Prediction of engine efficiency as a function of

compression ratio

Evaluation of power output with respect to

load

Effect of valve timing

Effect of speed on engine performance

STEP BY STEP APPROACH

As the first approach, ideal cycle simulation with

air as the working medium is undertaken. Ideal

cycle simulation is based on the following

assumptions viz. the working fluid throughout the

cycle is a fixed mass of air and is assumed to be an

ideal gas. There is no intake or exhaust process.

The combustion process is replaced by a heat

addition process from an external source; the heat

addition takes place at constant volume and is

assumed to be instantaneous; the cycle is

completed by heat rejection to the surroundings (in

contrast to the exhaust and intake processes of

actual engine); all processes are internally

reversible; there is no heat transfer to the

surroundings; there is no friction involved and the

working medium has constant specific heats (Cp

and Cv).

78



The main idea of SI engine simulation using ideal

Otto cycle is to enable the designers to examine

qualitatively the influence of the different variables

on its performance. The results obtained from ideal

cycle simulation, such as pressures, temperatures,

efficiency and mean effective pressure, will differ a

great deal from those pertaining to the actual

engine. The emphasis, therefore, in consideration

of the simulation of an ideal cycle will be primarily

on the qualitative aspects. To this extent the aim is

to calculate pressure, p, volume V and temperature

T at various points and also the work output and

thermal efficiency. Equations are available in

reference [1]. The important point to note is that

the efficiency of the ideal Otto cycle is a direct

function of compression ratio.

It is also true of an actual spark ignition engine that

the efficiency can be increased by increasing the

compression ratio. The trend towards higher

compression ratios is prompted by the effort to

obtain higher thermal efficiency. However, in the

actual engine there is an increased tendency

towards knocking as the compression ratio is

increased. Therefore, the maximum compression

ratio that can be used is fixed by the fact that the

knocking limit is not reached.

Fig.2 Block diagram depicting the details of

engine simulation

THERMODYNAMIC ASPECTS

Combustion in SI engines can be simulated ideally

under constant volume. Constant volume adiabatic

flame temperature calculation can be incorporated

to evaluate the peak temperature of combustion.

However, it may be noted that in an actual engine

the combustion is progressive, meaning

combustion spreads over a period of crank angle

(time). A typical GUI for calculating the adiabatic

flame temperature is shown in Fig.3.

The combustion of fuel and air mixture inside the

engine cylinders takes finite time interval. The

combustion duration and the rate of heat release

during combustion are most important and they

critically affect the peak pressure and temperature

thereby the engine performance. The GUI for

adiabatic flame temperature calculation for various

fuels, viz. gasoline, diesel, methanol, ethanol etc.

and operating conditions, such as full throttle and

part throttle, with either both reactant and product

in gaseous or liquid state is given in fig.3.

Appropriate equations are used in developing the

GUI [1]. GUI for ideal cycle simulation and fuel-

air cycle simulation is shown in Figs.4 and 5.

Progressive Combustion Analysis

In progressive combustion analysis, combustion is

assumed to spread over a period of crank angle

(time). In an actual engine, the combustion of fuel

and air mixture inside the engine cylinders takes

finite time interval and is progressive. The

combustion duration and the rate of heat release

during combustion are most important and they

critically affect the engine performance.

Ignition in SI engines starts before the piston

reaches top dead centre, and combustion is not

completed until the piston has moved beyond top

dead centre. It is evident that due to progressive

burning, the work done on the gases during the

compression process will be increased while the

work done by the gases during the expansion

process will be decreased, which will definitely

decrease the brake power output. The main

objective is to develop equation to calculate the

pressure-volume variation during the combustion

process and compute the power output when

progressive combustion is included in the analysis

and also to examine whether a reasonably

quantitative prediction with respect to the actual

engine could be obtained or not. The question is

how to represent the combustion process

mathematically.

79



Fig.3 GUI for Adiabatic Flame Temperature Calculation

Fig.4 GUI for Ideal Cycle Analysis

80



Fig.5 GUI for Progressive Combustion Analysis

If the combustion proceeds at some volume V, then

the pressure change ∆p is given by [1].

nV

Vpp

V

Vpkp td c

23

(1)

Equation (1) means that the pressure change during

combustion is due to two factors, viz. piston

movement and combustion. The first term in Eq.(1)

takes care of pressure change due to piston

movement whereas the second term is due to the

progressive combustion. The change in pressure is

related to volume change and mass fraction of gas.

Integration is carried out to obtain p and V values

during the progressive combustion process.

For progressive combustion analysis, three models

are used, viz.(1) Uniform rate model (2) Square

law model (3) Cosine law model.


To start with, the P-V diagram is predicted with

progressive combustion cycle and the result is

compared with the experimental result available in

[6].

VALIDATION OF THE PREDICTED

RESULTS

For the engine performance, first of all it is

necessary to evaluate to obtain the pressure-crank

angle diagram. From the P-V diagram, peak

pressure value and the crank angle at which the

peak pressure occurs can be determined.

Temperature can be determined at different crank

angles. Burning rate can also be determined from

the pressure values. A comparison has been made

between the experimental values and numerical

predictions (Fig. 6) for a typical engine given in

reference [6]. Predicted results are reasonably in

good agreement with the experimental values.

Fig. 6: Pressure Vs crank angle diagram – A

comparison between prediction and experiment

BRAKE POWER – Brake power is calculated from

brake mean effective pressure (which is obtained

from net work output) taking into account friction

mean effective pressure. The various frictional

losses are taken into account by appropriate

equations from the reference [12] from which bp

can be calculated. Predicted values of brake power

81



are compared with experimental values and are in

satisfactory agreement with each other.

Fig7: Comparison between experimental

values and numerical predictions on brake

power at 1000 rpm

Comparison between experimental and numerical

values is shown in Fig. 7. It can be seen from the

graph that the brake power increases with inlet

manifold pressure. Cylinder charge density

increases with inlet manifold pressure. So the

power output increases more or less linearly as the

inlet manifold pressure.

PARAMETRIC STUDIES ON ENGINE

PERFORMANCE

The effect of individual design variables on engine

performance can be studied with the help of cycle

simulation programs. Engine geometric variables

like compression ratio, cylinder bore, stroke,

connecting rod length, exhaust and intake valve

area as well as engine operating variables like

engine speed, air-fuel ratio and load affect the

engine processes like combustion, intake and

exhaust flow, heat transfer, friction etc. Thus the

required performance of engine can be obtained by

optimising the design variables. The effect of

compression ratio on brake thermal and mechanical

efficiency has been studied.

COMPRESSION RATIO – In actual engine other

processes like heat transfer, friction, and

combustion also play a part with the increasing

compression ratio. Figure 8 shows predicted results

of brake thermal efficiency for various

compression ratios. The increase in efficiency is

due to reduced exhaust gas dilution. Hence higher

compression ratios produce higher thermal

efficiencies. However the rate of increase reduces

with the increase in compression ratio.

Mechanical efficiency is one of the important

parameters in determining the performance of an

engine. As IP and BP are calculated using FP

relations, mechanical efficiency is calculated as the

ratio of BP to IP. Figure 9 shows variation of

mechanical efficiency against compression ratio.

Due to increase in compression ratio, brake mean

effective pressure increases. So, initially,

mechanical efficiency increases. At higher

compression ratio increased friction causes the rate

of increase of mechanical efficiency to decrease.

Fig.8 Effect of compression ratio on brake thermal

efficiency

CONCLUSION

From the present study it is concluded that simple

thermodynamic models can be effectively used to

study the performance of the engines to evaluate

the effect of various operating parameters. Out of

the various operating parameters it is noted that the

combustion model is the most important parameter

to be looked into more closely. Further,

progressive combustion analysis provides better

insight about the performance of the engine. It

could be seen that from the simulation model

developed and the validation of the predicted

results, it is possible to analyze the engine

82



performance in a much quicker and faster way.

Thus it is a useful tool for the researchers and the

practicing engineers.

Fig.9 Effect of compression ratio on

mechanical efficiency

REFERENCES

[1] Ganesan, V. (1996), Computer Simulation of

Spark-ignition Engine Processes,

Universities Press (India) Ltd. India

[2] Ganesan, V, 2000. Computer simulation of

Compression ignition engine processes,

University Press (India) Ltd. India

[3] Hayes, I. (ed.), Specification case studies.

Prentice-Hall 1987.

[4] Craigen, D., S. Gerhart, T. Ralston, Formal

methods reality check: Industrial usage. In:

F. C. P. Woodcock, P. G. Larsen (eds),

FME‘ 93, Lecture Notes in Computer

Science Vol. 670, Springer 1993, pp. 250-

267

[5] Bowen, J., Stavridou, V., The industrial

take-up of formal methods in safety critical

and other areas: A perspective. In: F. C. P.

Woodcock, P. G. Larsen (eds), FME‘ 93,

Lecture Notes in Computer Science Vol.

670, Springer 1993, pp. 183-195

[6] Lakshminarasimhan, V. (1993), Combustion

Modelling and Performance Simulation of

Four-stroke Spark-ignition Engine, M.S.

thesis, Indian Institute of Technology,

Madras, India.

[7] Mathur, H. B., Gajendra Babu, M. K. and

Subba Reddy, K., (1983), A

Thermodynamic Simulation of Model for a

Methanol Fuelled Spark-ignition Engine,

SAE Paper 831697

[8] Heywood, J.B, 1988. Fundamentals of

Internal Combustion Engine. McGraw Hill

Ltd, New York.

[9] Ball, J., Raine, R., and Stone R., (1998),

Combustion Analysis and Cycle-by-Cycle

Variation in Spark Ignition Engine

Combustion, Parts I and II, Proc. I. Mech.

E., Part D, Vol.212, Journal of Automotive

Engineering, London.

[10] Ball, J. K., Stone, C. R., Collings N. (1999),

Cycle by Cycle Modelling of NO Formation

and Comparison with Experimental Data, I.

Mech. E., Part D,213, J. Automotive Engg,

London.

[11] Bazari, Z., Smith, L. A., Banisoleiman, K.

and French, B.A.,(1996), An Engineering

Building Block Approach to Engine

Simulation with Special Reference to New

Application Areas, Paper C499/017,

Computers in Reciprocating Engines, Mech.

E. Conf. Publication, MEP, London.

[12] Bishop, I. N., (1964), Effect of Design

Variables on Friction and Economy, SAE

Paper 812-A

CONTACT

1, 2 IC Engines Division, Department of Mechanical

Engineering, Anna University, Chennai 600 025,

India

3 Internal Combustion Engines Laboratory

Department of Mechanical Engineering

Indian Institute of Technology Madras

Chennai 600 036, India

Email: [email protected]

83



N-2010-S01-23

Application of Statistical Tools in Evaluation of

Combustion Quality of SI Engines under Idling

Joginder Singh Kaliravna, Mangesh Nimbalkar, Narendra Kumar Jain

Tata Motors Ltd, Pimpri, Pune

Ganesan V

Internal Combustion Engines Laboratory, Department of Mechanical Engineering, Indian Institute of

Technology Madras, Chennai 600 036, India

ABSTRACT

In this work combustion quality of a bi-fuel SI

engine under idling condition is evaluated for two

different camshafts using statistical tools.

Statistical tools help us to define the customer‘s

subjective feeling for engine. Through statistical

tools it is easy to interpret that what were the

values of certain parameters when customers rated

that particular engine as good, average or bad.

Some of these tools used here are standard

deviation (SD), coefficient of variance (COV) and

lowest normalized value (Lnv). The standard

deviation (SD) and lowest normalized value (LNV)

of indicative mean effective pressure (IMEP) are

determined through the analysis of in-cylinder

pressure data of the engine. The parameter SD of

IMEP is a measure of the roughness or

unsteadiness of combustion and LNV is a measure

of the misfire tendency. Measurements of in-

cylinder pressure data recorded on engine with two

different camshafts in gasoline and CNG mode is

used to compute values of SD and LNV of IMEP.

Second camshaft has less overlap duration (in

terms of crank angle) and also more overlap area

than first camshaft. It has been observed that with

second camshaft value of SD and LNV of IMEP is

improved significantly which in turn shows

improved combustion stability at idling in both fuel

modes

INTRODUCTION

Now a days trend within engine testing is aimed at

reducing the time required for the development of

engine. During testing, data for large number of

variables related to engine can be acquired at the

same instant through data acquisition unit. For

proper analysis of this data we need certain

statistical tools. Statistical tools enable us to

understand data through summarized values and

graphical presentations. Summarized values not

only include the average, but also standard

deviation (SD), coefficient of variation (COV) and

lowest normalized value (LNV). It is important to

look at statistics along with the data set to

understand the entire picture Also in some cases

statistical tools help us to define the customer‘s

subjective feeling for engine. Through statistical

tools it is easy to interpret that what were the

values of certain parameters when customers rated

that particular engine as good, average or bad.

Some of these tools used here are standard

deviation (SD), coefficient of variance (COV) and

lowest normalized value (LVN). To investigate the

combustion stability at idling in gasoline and CNG

mode, the parameters such as SD and LNV of

IMEP introduced by Hoard and Rehagen [1] are

determined through the analysis of in-cylinder

pressure data of the engine. Fiorenza et al [2]

proposed a correlation between a parameter called

―cam overlap total area‖ and combustion

irregularity. Authors proposed two statistical

indices: (a) Standard deviation of IMEP, measure

of the instability of combustion. (b) Worst

Combustion Cycle (WCC) measure of tendency

toward misfiring. In order to obtain a good

combustion quality at idle, the standard deviation

was targeted to be less than 12 kPa, while WCC

was targeted to be at least 70%. Johan et al [3]

investigated the influence of valve overlap

strategies on the residual gas fraction, combustion

parameters and cycle to cycle IMEP variations in

84



SI engine at idling. The aim of this study is to

investigate the effect of cam profile and valve

events on combustion stability under idling

condition. In second camshaft, cam profile and

valve events are optimized to reduce the valve

overlap duration and total valve overlap area. This

leads to less residual gas contamination in inlet

manifold during idling, which in turn may improve

the combustion stability in both fuel modes. The

residual gas contamination in inlet manifold

depends on following parameters. (a) The pressure

difference between inlet manifold and exhaust

manifold. (b) The valve opening duration. (c)

Valve opening area during overlap. At idling

condition, throttle valve is almost at zero position

hence causes maximum pressure difference

(approx. 0.6bar) between intake and exhaust

manifold. This is the driving force for flow of

residual gases from combustion chamber to intake

port. The pressure difference is due to throttling in

inlet and back pressure in exhaust system. It is

usually fixed because of idling conditions; hence

pressure difference cannot be reduced notably.

However, the valve overlap duration in terms of

crank angle degree and the total valve opening area

during overlap are the parameters; these can be

used to reduce residual gas contamination

considerably in intake manifold and improves

combustion stability.

CAMSHAFT PARAMETERS

Camshafts named as first and second having

different cam profile and valve timings are

evaluated through measurements. The effects of

following camshaft parameters were studied:

1. Valve lift curve for exhaust and intake

valves

2. Valve overlap area

3. Valve events (IVO, IVC, EVO and EVC)

VALVE LIFT CURVE FOR EXHAUST AND

INTAKE VALVES - Figure 1 shows normalized

valve lift curves for intake and exhaust valves with

respect to crank angle degree for both camshafts.

Valve lift curves have been normalized with

respect to maximum lift of exhaust valve of second

camshaft. TDC at the start of power stroke is taken

as reference (i.e. zero crank angle degree) for

plotting valve lift curves. This is done so to show

exhaust and intake valve overlap (i.e. from intake

valve opening to exhaust valve closing) duration.

Figure 1: Normalized valve lift curves

It can be seen from Fig. 1 that in second camshaft,

ramp of exhaust and intake lift have been made

steeper to reduce effective valve overlap area. Also

in second camshaft, maximum exhaust lift is

increased by 0.3 mm and maximum intake lift is

increased by 0.55 mm.

VALVE OVERLAP AREA – Figure 2 shows the

comparison of valve overlap area from intake valve

opening to exhaust valve closing for both

camshafts. Crank angle is plotted from 270 deg to

540 deg to show enlarged view of valve overlap

duration. Zero degree crank angle represents TDC

of the power stroke and 360 degree crank angle

represents start of suction stroke for next cycle.

Effectively there is 40% reduction in valve overlap

area from first camshaft to second camshaft. Less

valve overlap area will help in reducing residual

gas contamination in the intake manifold under

idling conditions, as in this condition, throttle valve

is almost at zero position, thereby intake manifold

and intake ports average pressure is even less than

atmospheric pressure (approx 0.46 bar absolute) on

the other hand in exhaust manifold and exhaust

ports average pressure is close to the atmospheric

pressure (approx 0.95 bar absolute) due to back

pressure created by catalytic converter. So the

85



maximum pressure difference between intake and

exhaust ports is approx. 0.6 bar in idling condition,

which is the driving force for flow of combustion

products from combustion chamber into the intake

port. Now as piston crosses TDC, to induct fresh

charge from intake port for next cycle, the residual

gases of the previous cycle which have gone into

the intake port will return back first into cylinder,

this can cause poor combustion in this cycle. Lesser

the residual gas contamination in intake port from

previous cycle will be better for next cycle; this in

turn will improve combustion stability at idling.

However in WOT conditions pressure difference

will be quite small between intake and exhaust

port, so relatively there will be low residual gas

contamination, which is good for combustion

stability in WOT condition.

VALVE EVENTS – As can be seen from Fig. 1

EVO, IVO, EVC and IVC events are different for

both camshafts. These events are defined at 0.1mm

lift of corresponding intake and exhaust valve as

shown in Table 1.

Table 1: Valve events (in terms of crank angle deg)

EVO (BBDC)

IVO (BTDC)

EVC (ATDC)

IVC (ABDC)

Overlap Duration

First

Camshaft 52 27 25 59 52

Second

Camshaft 47 15 13 45 28

Figure 2: Valve opening area during overlap

EXPERIMENTAL SETUP

The engine under consideration is mounted on a

frame using rubber pads as in vehicle condition to

reduce the transmission of vibrations. For full load

performance, engine crankshaft is connected to the

dynamometer. For idling performance, engine has

to be disconnected from dynamometer. A heat

exchanger is used for cooling engine coolant.

Coolant and lubrication oil temperature is

maintained at 90±3 deg C. Wiring harness is fitted

and each connector is connected to corresponding

sensor or actuator along with a 12-volt battery.

Throttle is controlled by separate servo motor.

Exhaust system is connected to the engine for

correct simulation of exhaust gas back pressure.

Engine is well instrumented to measure all

important testing parameters. Crank angle based

intake and exhaust port pressures in combination

with in-cylinder pressure are measured. For

calculating in-cylinder and exhaust pressures water

cooled pressure sensors are used. AVL-Indimaster

synchronized with optical crank angle encoder is

used for acquiring pressure data measured by

pressure sensors. Representation of instruments and

engine test set-up is shown in Fig.3.

Figure 3: Representation of instrumented engine

on test bed

EXPERIMENTAL RESULTS

Measurements of in-cylinder pressure with respect

to crank angle are done to investigate the variation

in combustion with both camshafts for gasoline and

CNG mode of operation.

IN-CYLINDER PRESSURE MEASUREMENT

WITH FIRST CAMSHAFT – Figure 4 shows in-

cylinder pressure data for gasoline and CNG mode

with first camshaft. It can be observed from figure

86



that at EVO, in-cylinder pressure is high in

gasoline mode (this will assist in blowdown)

compared to CNG mode. Further it can be seen that

the magnitude of in-cylinder pressure after EVO in

gasoline mode has fallen more than the CNG mode.

Which means less amount of residual content will

flow from combustion chamber to intake manifold

in case of gasoline mode as compared to CNG

mode under idling condition. This may lead to

better combustion in gasoline mode as compared to

CNG mode. Measured data of IMEP also shows

that gasoline has better combustion stability than

CNG presented in subsequent sections.

Figure 4: Measured in-cylinder pressure with First

camshaft at idling

IN-CYLINDER PRESSURE MEASUREMENT

WITH SECOND CAMSHAFT – Figure 5 shows

in-cylinder pressure data for gasoline and CNG

modes with second camshaft. It can be observed

that at EVO, in-cylinder pressure is high in both

fuel modes as compared to first camshaft (refer Fig.

4), which will assist in blow-down and better

scavenging. Also because of same in-cylinder

pressure at EVO in both fuel modes the pressure

curve is of same magnitude with second camshaft

which is not the case with first camshaft.

MEASURED IMEP VARIATONS – Using AVL-

Indimaster, in-cylinder pressure data is recorded at

an interval of 0.5 deg crank angle. IMEP values

have been calculated in AVL-Indimaster for 200

consecutive cycles in gasoline and CNG mode for

both camshafts.

Figure 5: Measured in-cylinder pressure in with

Second camshaft at idling

MEASURED IMEP VARIATION WITH FIRST

CAMSHAFT IN GASOLINE MODE – Figure 6

shows IMEP values of 200 consecutive cycles for

gasoline mode with first camshaft. The average

value of IMEP is 1.84 and standard deviation of

IMEP is 0.69.

Figure 6: Measured IMEP variation for 200 cycles

in gasoline mode with first camshaft

IMEP Variation for 200 cycles in gasoline mode

with base camshaft at idling

-1

-0.5

0

0.5

1

1.5

2

2.5

3

3.5

4

0 50 100 150 200Cycles

IME

P

IMEP

avg IMEPavg IMEP + SD IMEP

avg IMEP - SD IMEP

EVO

EVO

87



MEASURED IMEP VARIATION WITH FIRST

CAMSHAFT IN CNG MODE – Figure 7 shows

IMEP values of 200 consecutive cycles in CNG

mode with first camshaft. The average value of

IMEP is 1.46 which is 20.6% less compared to

gasoline mode (refer Fig. 6). Standard deviation of

IMEP is 1.07 (this is because of larger variation in

IMEP values from average IMEP). This can be

attributed to more residual gas contamination of

fresh charge in CNG mode.


in CNG mode with first camshaft

MEASURED IMEP VARIATION WITH

SECOND CAMSHAFT IN GASOLINE MODE –

Figure 8 shows IMEP values of 200 consecutive

cycles for gasoline mode with second camshaft.

The average value of IMEP is 1.7 which is 7% less

compared to gasoline mode with first camshaft

(refer Fig. 6).


in gasoline mode with second camshaft

The standard deviation of IMEP is 0.35. This

shows that the variation in IMEP values from

cycle-to-cycle is reduced by 49% with second

camshaft as compared to first camshaft, which in

turn may lead to improved combustion stability

under idling condition.

MEASURED IMEP VARIATION WITH

SECOND CAMSHAFT IN CNG MODE – Figure

9 shows IMEP values of 200 consecutive cycles for

CNG mode with second camshaft. The average

value of IMEP is 1.52 which is 4% more compared

to CNG mode with first camshaft (refer Fig. 7).


in CNG mode with second camshaft

Table 2: comparison of SD, COV and LNV values

of IMEP

IMEP

CNG Mode Gasoline Mode

First

camsh

aft

Secon

d

camsh

aft

Impro

ve

ment

(%)

First

camsh

aft

Secon

d

camsh

aft

Impro

ve

ment

(%)

SD 1.07 0.44 59 0.69 0.35 49

Minimu

m -0.076 0.66 0.48 0.79

Averag

e 1.46 1.52 1.84 1.69

COV

(%) 73.68 28.84 61 37.58 20.81 45

LNV

(%) -5.39 43.29 112 25.88 47.11 45

IMEP Variation for 200 cycles in CNG mode

with base camshaft at idling

-1

-0.5

0

0.5

1

1.5

2

2.5

3

3.5

4

0 50 100 150 200Cycles

IME

P

IMEP


avg IMEP - SD IMEP

IMEP Variation for 200 cycles in gasoline mode

with new camshaft at idling

-1

-0.5

0

0.5

1

1.5

2

2.5

3

3.5

4

0 50 100 150 200Cycles

IME

P

IMEP


avg IMEP - SD IMEP

IMEP Variation for 200 cycles in CNG mode

with new camshaft at idling

-1

-0.5

0

0.5

1

1.5

2

2.5

3

3.5

4

0 50 100 150 200Cycles

IME

P

IMEP


avg IMEP - SD IMEP

88



The standard deviation of IMEP is 0.44. This

shows variation in IMEP values from cycle-to-

cycle is reduced by 59% with second camshaft as

compared to first camshaft in CNG mode. Hence

with the second camshaft there is significant

improvement while operating under CNG mode

compared to first camshaft. This can also be

observed from Fig. 5 (i.e. at EVO higher in-

cylinder pressure assisted in better scavenging and

less residual gas contamination in inlet manifold).

SD, COV AND LNV OF IMEP – Table 2 shows

the comparison of standard deviation (SD),

coefficient of variance (COV) and lowest

normalized value (LNV) in gasoline and CNG

mode for both camshafts.

CONCLUSION

Based on the experiments carried it is concluded

that for the engine under consideration the

combustion stability under idling condition has

improved significantly in both fuel modes by

reducing the valve overlap area by 40%. From this

study it can also be concluded that the Statistical

tools are very useful in evaluation of combustion

quality

REFERENCES

John Hoard and Rehagen, ―Relating Subjective Idle

Quality to Engine combustion‖, SAE Paper

No, 970035

R. Fiorenza, G. Formisano and F. Petraglia, ―A

Calculation Methodology for Cam Overlap

Optimization towards Combustion Quality at

Idle in IC SI Engines‖, SAE Paper No. 2003-

32-0040

Hakan Sandquist and Johan Wallesten, ―Influence

of valve overlap strategies on residual gas

fraction and combustion in a spark ignition

engine at idle‖, SAE Paper No. 972936

89



N-2010-E05-24

Experimental investigation on PCCI combustion in a

single Cylinder DI engine

1 K. Bhaskar,

2 R. Ameerkhan,

2 M. Dinesh,

3 Dr. G. Nagarajan

4

Keywords: HCCI emission, IC engine,

ABSTRACT

The present investigation was to study the

performance, combustion and emission

characteristics of homogeneous charge

compression ignition (HCCI) combustion of diesel

fuel with external mixture formation technique. A

stationary four stroke, single cylinder, direct

injection diesel engine capable of developing 4.4

kW at 1500 rpm was modified to operate in HCCI

mode. To achieve homogeneous mixture, diesel

fuel was injected in the intake manifold by using a

solenoid operated injector with pressure regulator.

To control the early ignition of diesel, cooled EGR

technique was adopted. Experiments were

conducted with manifold injection without EGR

and manifold injection with 10% EGR and results

are compared with conventional diesel fuel

operation (DI @ 23 deg bTDC and 200 bar

injection pressure). From the experimental results,

it is found that, the ignition delay is reduced

considerably for manifold injection due to better

mixture preparation and results in low emissions. A

reduction of about 55% and 80% in NOx emissions

and 20% and 30% reduction in smoke emission are

obtained for manifold injection without EGR and

manifold injection with 10% EGR compared to

conventional mode of operation.

INTRODUCTION

In the quest for ever improving fuel efficiency and

emissions reduction, an old and very promising

idea has found new life. HCCI homogeneous

charge technology has been around for a long time,

but has recently received renewed attention and

enthusiasm. While the early years saw many

insurmountable obstacles whose answers would

only come as sophisticated computer controlled

electronics were developed and matured into

reliable technologies, progress stalled. Time has, as

it always does, worked its magic and nearly every

problem has been solved. HCCI is an idea whose

time has come with nearly all of the parts and

pieces of technology and know-how in place to

make a real go of it.

Before we attempted to begin our process of

experimental analysis of the HCCI combustion

mode, we made a thorough study of the previous

attempts that were carried out, which gave us a

deep insight of the various modifications which

ought to be carried out to make the HCCI

combustion feasible.

Ganesh et al investigated the single cylinder diesel

engine to run on a Homogenous fuel/air mixture

that is generated externally in a fuel Vaporizer.

They further reported that low NOx and smoke

emission was achieved with homogenized mixture.

Najt et al investigated effect of exhaust emission in

the Homogeneous charge diesel combustion

(HCDC) in a diesel engine. They further reported

the effect of supplying pre-mixed fuel into the

intake manifold.

EXPERIMENTAL SETUP

An experimental setup has been developed to

conduct test on four stroke, single cylinder,

vertical, air-cooled, diesel engine. Necessary

instruments were provided after inspection and

calibration to evaluate performance, emissions and

engine parameters at different operating conditions.

The schematic of experimental setup is shown in

the Fig. 1 and Table 1 shows the test engine

specifications.

90



Figure 1 Experimental Setup

1 Diesel Engine 9 TDC position

sensor

2 Electrical

Dynamometer

10 Charge amplifier

3 Dynamometer

Controls

11 TDC amplifier

circuit

4 Air Box 12 A/D Card

5 U tube manometer 13 Personal computer

6 Fuel Tank 14 Exhaust gas

analyzers

7 Fuel measurement 15 AVL smoke meter

8 Pressure pickup

Table 1: Test Engine Specifications

Engine Type

Four stroke, Air cooled,

stationary, constant speed,

direct injection, CI engine

No. of cylinders 1

Maximum power 4.4 kW at 1500 rpm

Maximum torque 28 N-m at 1500 rpm

Bore 87.5 mm

Stroke 110 mm

Displacement 661.5cc

Compression Ratio 17.5: 1

Injection Timing 23.40 bTDC

Loading type Swinging field dynamometer

MANIFOLD FUEL INJECTION SET UP

The fuel is injected in the inlet manifold using an

electronic fuel pump through a fuel injector and the

fuel line pressure range is 0-6 bars, according to

ratio of pilot injection quantity of fuel to the main

injection. The rating of the fuel injection pump that

we have used is 20 ampere at 12 volts and that of

the fuel injector is 0.3 ampere.

FUEL INJECTION CONTROL UNIT

The injection is controlled by electronic circuit

having a limit switch with frequency of about 750

cycles per minute. The limit switch is actuated

from the inlet valve rocker having a 9mm travel.

This is used for initiating the starting and ending of

injection during the suction stroke with the desired

pressure by solenoid actuated injector and pressure

can be controlled by pressure regulator valve for

different loading conditions. Fig. 2 shows

schematic diagram of manifold fuel injection.

Figure 2 schematic diagram of manifold fuel

injection system

EGR SET UP

The cooled EGR setup is fabricated in the intake

system which allows exhaust gas recirculation into

intake manifold. EGR percentage can be varied

with the help of control valves provided in the

setup. Fig. 3 shows the schematic layout of cooled

EGR set up.

91



Figure 3 schematic layout of cooled EGR set up

EXPERIMENTAL PROCEDURE

The engine was started with diesel and allowed to

warm up to attain the steady state condition. Engine

Speed, fuel consumption rate, Exhaust Gas

Analysis (HC, CO, CO2, NOX and ) using MRU 5

Gas Analyzer, Smoke Intensity using AVL Smoke

meter, pressure-crank angle diagram with AVL

‗Indimeter‘ software and exhaust gas temperature

were measured at various loads. The experiment is

repeated with manifold injection with EGR and

without EGR.

The Premixed ratio (Rp) is defined as the ratio of

energy of premixed fuel Qp to total energy Qt. The

premixed ratio can be obtained from the following

equation.

Where,

mp is the mass of premixed fuel,

md is the mass of directly injected fuel,

hu is the lower heating value, and

Subscripts ‗p‘ and ‗d‘ are the premixed and directly

injected fuel, respectively.

The accurate measurement of EGR rate is the

premise to control EGR, but it is difficult by the

present-days technology. There are two common

measurements of EGR rate: 1) concentrations of

CO2 in intake and output gas and (2) air/fuel ratio.

The first method was used in this investigation, and

the formula used was as follows:

EGR % = ((CO2 %) intake / (CO2 %) exhaust)*100%

EGR is introduced into intake manifold by opening

EGR control valve and measured the value of CO2

in the intake manifold. Then the readings were

taken for all loads.


Experiments were conducted with 25 % Rp (PCCI

mode) and with 25 % Rp and 10 % EGR in the

manifold and the results were compared with base

line readings.

BRAKE THERMAL EFFICIENCY

The variation of brake thermal efficiency with load

for all the modes of operation is shown in Fig.4, It

can be seen that the brake thermal efficiency

decreases with the Rp and increases with EGR. At

any load the brake thermal efficiency decreases

with manifold injection due to poor vaporization of

the fuel in the manifold and due to wall wetting.

With EGR vaporization in the manifold is

improved and efficiency is better than that with

manifold injection.

COMBUSTION CHARACTERISTICS

Figure 5 shows the pressure-crank angle diagram

for conventional mode of diesel operation and

PCCI-DI without EGR and with 10% EGR at full

load. Form the figure it is observed that combustion

starts earlier than that of diesel for manifold

injection with and without EGR. It can be seen that

the peak pressure with manifold injection is the

highest without EGR and with EGR slightly less

but higher than diesel. The Peak pressure occurs

later in the expansion stroke with manifold

injection and it occurs earlier compared to manifold

injection with EGR. With EGR peak pressure is

higher than that of diesel and also occurs later in

the expansion stroke.

92



Figure 4 Brake thermal efficiency Vs Brake

Power

Figure 5 Pressure Vs Crank Angle

OXIDES OF NITROGEN EMISSIONS

Figure 6 shows the emission of oxides of nitrogen

for the three modes of operations considered. With

PCCI mode combustion occurs more or less

simultaneously in the entire cylinder. So the high

temperature region and high concentration of fuel

in the cylinder are eliminated which results in

lower NOx emissions. Induction of EGR with

premixed charge induction results in further

reduction in oxides of nitrogen due to reduction in

combustion temperature.

SMOKE EMISSIONS

The variation of smoke with load for the three

modes of operation is shown in figure 7. With

manifold injection the smoke level is the lowest at

all loads when compared to conventional diesel

fuel operation due to the disappearance of rich

regions of mixture inside the combustion chamber.

Smoke level slightly increases with the induction

of EGR due to reduction in availability of oxygen

for combustion.

Figure 6 Oxides of Nitrogen (NOx) Vs Brake

Power

Figure 7 Smoke Vs Brake Power

HYDROCARBON EMISSIONS

Figure 8 shows the HC emission at various loads

for the three modes. It can be seen that HC

emissions are high at all loads for the PCCI mode.

Induction of EGR increases the HC emission

further due to lower peak temperature and

reduction of oxygen content.

Figure 9 shows the CO emission at various load for

three the modes. It can be seen that CO emissions

are low up to 75% of load. PCCI results slightly

higher CO emissions due to the low temperature

0

5

10

15

20

25

30

0 1.1 2.2 3.3 4.4

Brake Power (kW)

Th

erm

al

Eff

icie

ncy (

%)

Diesel mode

Rp 25 %

Rp 25 % 10 EGR

0

10

20

30

40

50

60

70

80

-30 -20 -10 0 10 20 30

Crank Anagle (oC)

Pre

ssu

re (

bar)

Diesel

Rp 25

Rp 25% 10% EGR

0

200

400

600

800

1000

1200

1400

1600

0 1.1 2.2 3.3 4.4

Brake Power (kW)N

Ox (

pp

m)

Diesel mode

Rp 25 %

Rp 25 % 10 EGR

0

20

40

60

80

100

120

140

160

180

0 1.1 2.2 3.3 4.4Brake Power (kW)

Sm

oke (

mg

/m3)

Diesel mode

Rp 25 %

Rp 25 % 10 EGR

93



combustion of lean mixture. With induction of

EGR CO emission is slightly increased at loads

compared to PCCI mode.

Figure 8 Hydrocarbon (HC) Vs Brake Power

CARBON MONOXIDE EMISSIONS

Figure 9 CO Vs Brake Power

Figure 10 Exhaust Gas Temperature Vs BP

EXHAUST GAS TEMPERATURE

The variation of exhaust gas temperature at various

loads for the three modes of operation is shown in

figure 10. It is observed that PCCI mode

combustion and PCCI mode with EGR results in

lowest exhaust gas temperature at all loads.

CONCLUSION

Based on the investigations carried out it can be

concluded that PCCI-DI mode effectively reduces

the NOx and Smoke emissions while HC and CO

emissions are slightly increased.

At all loads the brake thermal efficiency decreases

with manifold injection due to poor vaporization of

the fuel in the manifold and due to wall wetting.

With EGR vaporization in the manifold is

improved and efficiency is better than that with

manifold injection. The decrease in efficiency can

be overcome by increasing inlet charge

temperature.

REFERENCES

1. D. Ganesh, G. Nagarajan and M. Mohamed

Ibrahim, ―Study of performance, combustion

and emission characteristics of diesel

homogeneous charge compression ignition

(HCCI) combustion with external mixture

formation‖ Fuel 87 (2008) 3497–3503.

2. P. Najt, D.E. Foster: ‖Compression-Ignited

Homogeneous Charge Combustion‖,

SAE830264

3. R.H. Thring, ‖Homogeneous Charge

Compression Ignition(HCCI) Engines‖,

SAE892068.

4. T. Aoyama, Y. Hattori, J. Mizuta, Y. Sato:

‖An Experimental Study on Premixed-Charge

Compression Ignition Gasoline Engine‖,

SAE960081

5. T.W. RYAN, T.J. CALLAHAN:

‖Homogeneous Charge Compression Ignition

of Diesel Fuel‖, SAE961160

CONTACT

1 Assistant Professor, Department of Automobile

Engineering, SVCE, Sriperumbudur, 602105.

Tamil Nadu 2, 3

Graduate students, Department of Automobile

Engineering, Sri Venkateswara College of

Engineering, Sriperumbudur, 602105, Tamil Nadu 4 Professor, Department of Mechanical

Engineering, IC Engine Division, College of

Engineering, Anna University Chennai, Chennai.

0

10

20

30

40

50

60

0 1.1 2.2 3.3 4.4

Brake Power (kW)

HC

(p

pm

)

Diesel mode

Rp 25 %

Rp 25 % 10 EGR

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 1.1 2.2 3.3 4.4

Brake Power (kW)

CO

(%

Vo

l.)

Diesel mode

Rp 25 %

Rp 25 % 10 EGR

100

150

200

250

300

350

400

450

500

0 1 2 3 4 5

Brake Power (kW)

Exh

au

st

Tem

pera

ture

(oC

)

Diesel mode

Rp 25 %

Rp 25 % 10 EGR

94





N-2010-D13-28

Experimental Study on Disc Brake Squeal

K. Saravanan

College of Engineering Guindy, Anna University Chennai, Tamil Nadu, INDIA

KEYWORDS: Disc brake, squeal, finite element , complex eigenvalue, parametric study.

ABSTRACT

Disc brake squeal noise is a very complicated

phenomenon, which automobile manufacturers

have confronted for decades, due to consistent

customer complaints and high warranty costs. In

recent years, the finite element method (FEM) has

become the preferred method due to high hardware

costs of experimental methods. In this study, a

simplified model for the disc brake is presented

using the Abaqus/Standard finite element software.

The analysis process uses a nonlinear static

simulation sequence followed by a complex Eigen

value extraction to determine the squeal propensity.

The effect of the main operational parameters

(braking pressure, and friction coefficient) on the

squeal propensity is performed. The influence of

changing the rotor stiffness and back plates

stiffness, under different operation conditions is

investigated. The results of this analysis show that

the squeal noise can be reduced by increasing the

rotor stiffness and decreasing the back plate

stiffness of the pads.

INTRODUCTION

Disc brake noise, in general, is one of the major

contributors to the automotive industry‘s warranty

costs. In most cases, this type of noise has little or

no effect on the performance of brake system.

However, most customers perceive this noise as a

problem and demand that their dealer‘s fix it.

Customer complaints result in significant yearly

warranty costs. More importantly, customer

dissatisfaction may result in the rejection of certain

brands of brake systems or vehicles. The

automotive industry is thus looking for new ways

to solve this problem [1].

In general, brake noise has been divided into three

categories, in relation to the frequency of noise

occurrence. The three categories presented are low

frequency noise, low-frequency squeal and high-

frequency squeal. Low-frequency disc brake noise

typically occurs in the frequency range between

100 and 1000 Hz. Typical noises that reside in this

category are grunt, groan, grind and moan. This

type of noise is caused by friction material

excitation at the rotor and lining interface. The

energy is transmitted as a vibratory response

through the brake corner and couples with other

chassis components [2].

Low-frequency squeal is generally classified as a

noise having a narrow frequency bandwidth in the

frequency range above 1000 Hz, but below the first

in plane mode of the rotor. The failure mode for

this category of squeal can be associated with

frictional excitation coupled with a phenomenon

referred to as ‗‗mode locking‘‘ of brake corner

components. Mode locking is the coupling of two

or more modes of various structures producing

optimum conditions for brake squeal [2].

High-frequency brake squeal is defined as a noise

which is produced by friction induced excitation

imparted by coupled resonances (closed spaced

modes) of the rotor itself as well as other brake

components. It is typically classified as squeal

noise occurring at frequencies above 5 kHz. Since

it is a range of frequency which affects a region of

high sensitivity in the human ear, high-frequency

brake squeal is considered the most annoying type

of noise. Brake squeal is a concern in the

automotive industry that has challenged many

researchers and engineers for years. Considerable

analytical, numerical and experimental efforts have

been spent on this subject, and much physical

insight has been gained on how disc brakes may

95



generate squeal, although all the mechanisms have

not been completely understood [3].

This study attempts to present a simplified finite

element model to examine the squeal propensity of

a disc brake system for a range of operational

parameters like friction coefficient, and braking

pressure. The evaluation of the effect of material

properties (the rotor young‘s modulus and the back

plates of the pads young‘s modulus) on the squeal

propensity is performed. The simulations

performed in this study present a guideline to

reduce squeal noise by using design modification,

which dependent on the modified material

prosperities of disc brake components.

BRAKE NOISE GENERATION

MECHANISMS

Disc brake squeal occurs when a system

experiences vibrations with very large mechanical

amplitude. It is supposed that there are two

occurrence mechanisms of a squeal noise. The first

mechanism is a phenomenon resulting from the

―stick-slip‖ of a friction side [4]. The second

mechanism is a phenomenon resulting from

geometric instabilities of the brake assembly [3].

Both mechanisms, however, attribute the brake

system vibration and the accompanying audible

noise to variable friction forces at the pad–rotor

interface. Regarding the squeal noise caused by

geometric instability of system, if two neighboring

vibration modes are close to each other in the

frequency range and have similar characteristics,

they may merge if the coefficient of friction

between the pad and disc increases. When these

modes coupled at the same frequency, one of them

becomes unstable. The unstable mode can be

identified during complex Eigen value analysis [5-

12] because the real part of the Eigen value

corresponding to an unstable mode is positive.

METHODOLOGY AND NUMERICAL

MODEL

Problem formulation

The mass matrix and stiffness matrix of

engineering structures can be assumed to be

symmetric, respectively, positive definite and semi-

positive definite in general. The Eigen solutions of

such structures are extensively studied and the

vibration of such systems is stable. There are,

however, engineering problems whose stiffness

matrices are asymmetric. Usually the asymmetry is

produced not by the structure itself, but by some

external loads interacting with the structure [15],

such as friction in brake noise problems [13].The

equation of motion for a vibrating system is

---- (1)

Where M, C and K are mass, damping and stiffness

matrices, respectively, and u is the generalized

displacement vector. For friction induced vibration,

it is assumed that the forcing function F is mainly

contributed to by the variable friction force at the

pad-rotor interface. The friction interface is

modelled as an array of friction springs.

With this simplified interface model, the force

vector becomes linear:

{F}= [Kf ]{u} ----(2)

Where, Kf is the friction stiffness matrix. A

homogeneous equation is the obtained

By combining equations (1) and (2) and by moving

the friction term to the left-hand side

---- (3)

Eq. (3) is now the equation of motion for a free

vibration system with a pseudo forcing function in

the stiffness term. The friction stiffness acts as the

so-called ‗‗direct current‘‘ spring [1] that causes

the stiffness matrix to be asymmetric.

Complex Eigen value analysis

The complex Eigen value analysis made available

in ABAQUS is utilized to determine disc brake

assembly stability. The essence of this method lies

in the asymmetric stiffness matrix that is derived

from the contact stiffness and the friction

coefficient at the disc/ pads interface [6]. In order

to perform the complex Eigen value analysis using

ABAQUS, four main steps are required [7].

They are given as follows:

96



1. Nonlinear static analysis for applying brake-

line pressure

2. Nonlinear static analysis to impose rotational

speed on the disc

3. Normal mode analysis to extract natural

frequency of un-damped system

4. Complex Eigen value analysis that

incorporates the effect of friction coupling

In this analysis, the complex Eigen problem is

solved using the subspace projection method; thus,

the natural frequency extraction analysis must be

performed first to determine the projection

subspace. The complex Eigen value problem can

be given in the following form:

(λ2M + λ C + K) y = 0 ---- (4)

Where M is the mass matrix, which is symmetric

and positive definite; C is the damping matrix,

which can include friction-induced damping effects

as well as material damping contribution; K is the

asymmetric (due to friction contributions) stiffness

matrix; λ is the Eigen value; and y is the

eigenvector. Both Eigen values and eigenvectors

may be complex. In the third step stated above, this

system is symmetrised by dropping the damping

matrix C and asymmetric contributions to the

stiffness matrix s K to find the projection subspace.

Therefore the Eigen value, λ becomes a pure

imaginary where λ = iω and the Eigen problem can

be written as follows:

(−ω 2M + K s) Z = 0 ---- (5)

This symmetric Eigen value problem is solved using

the Lanczos iteration Eigen solver. Next, the

original matrices are projected onto the subspace of

real Eigen vectors z and given as follows:

M * = [ z1,z2 ,.... zn]

T M[[ z1,z2 ,.... zn], ----(6)

C * = [ z1,z2 ,.....zn]

T C[ z1,z2 ,...... zn], ----(7)

K * = [ z1,z2 ,.... zn]

T K[ z1,z2 ,....... zn], ----(8)

Now the Eigen value problem is expressed in the

following form:

(λ2M

* +λC

* + K

*) y

* = 0 ---- (9)

The reduced complex Eigen values problem is then

solved using the QZ method for a generalized non-

symmetrical Eigen value problem. The

eigenvectors of the original system are recovered

by the following:

Yk = [z1, z2,…………..zn] y

*k ----(10)

Where Yk is the approximation of the k-th

eigenvector of the original system. For more

detailed description of the formulation and the

algorithm we refer to [16].

The complex values λ, can be expressed as λ = α ±

iω where α is the damping coefficient (real part of

λ) and ω is the damped natural frequency

(imaginary part of λ) describing damped sinusoidal

motion. If the damping coefficient is negative,

decaying oscillations typical of a stable system

result. A positive damping coefficient, however,

causes the amplitude of oscillations to increase

with time. Therefore the system is not stable when

the damping coefficient is positive. By examining

the real part of the system Eigen values the modes

that are unstable and likely to produce squeal are

revealed. An extra term, damping ratio, is defined

as − 2α /ω .If the damping ratio is negative; the

system becomes unstable, and vice versa.

ANALYSIS OF STABILITY FOR DISC

BRAKE

Description of unstable modes of disc brake

To demonstrate the squeal propensity of the disc

brake, the 100 Eigen values extracted between zero

and 13 kHz for the base brake system with μ=0.5

are plotted on the complex plane in Figure 1. In the

baseline case no other sources of damping are

specified. All of the modes have zero damping (lie

on the imaginary axis) except where pairs of modes

have become coupled and formed a stable/unstable

pair. These result in the Eigen value that occurs in

conjugate pairs which are symmetrically located

about the imaginary axis. In this case nine unstable

modes can be seen. An alternative way to express

these results is to plot damping ratio vs. frequency

as shown in Figure 2. The nine modes with positive

real parts now appear with negative damping

values. While there is no direct proportionality

between squeal propensity and the level of

97



damping coefficient, it has been suggested that

higher values tend to be associated with modes that

are most likely to squeal [6].

Figure 1: Eigen values extracted from the disc

brake model plotted on the complex Plane

Figure 2: Damping ratio vs. frequency for the

disc brake model

The results show that higher damping coefficient is

approximately at unstable frequency 12 kHz. There

is a significant pad bending vibration for these

cases.

EFFECT OF PARAMETERS FOR DISC

BRAKE SQUEAL

Variation of friction coefficient

The effect of friction coefficient of the pad-rotor

interface is performed. Usually, the analysis is

performed for varying the friction coefficients from

0.1 to 0.7. With the low friction coefficient all of

the modes of the system will be stable. As the

friction coefficient is increased, modes can be

driven closer to one another in frequency. At some

critical friction value, a sudden change occurs

(called a bifurcation), and a new mode exists that

contains the original modes as a coupled pair.

Figure 3a, shows results in the form of the damping

coefficient as a function of frequency for different

friction coefficients. It can be seen that the major

squeal frequency is approximately 12 kHz. The

value of the damping coefficient increases

significantly with an increase of the friction

coefficient as shown in Figure 3b, at a frequency of

12 kHz.

It is understandable that with an increase in the

friction coefficient, there is an accompanying

increase in the instability of the system, thus an

increase in the damping coefficient. This means

that the most fundamental method of eliminating

brake squeal is to reduce the friction between the

pads and the disc. However, this obviously reduces

braking performance and is not a preferable method

to employ.

Figure 3a: Unstable modes with friction

coefficient varied from 0.1 to 0.7.

Variation of stiffness of the disc

The effect of rotor stiffness in terms of Young‘s

Modulus is performed. The rotor is made of grey

cast iron. The elastic modulus of cast irons varies

from below 100 GPa through to the values close to

that of steel at approximately 200 GPa. Grey cast

iron is particularly variable in properties depending

upon its carbon and, to a lesser degree, silicon

content [14]. The stiffness of the disc brake is

performed by varying Young‘s modulus of the disc

98



from 85GPa to 135GPa, where the baseline

young‘s modulus of the disc is 105 GPa. Figure 4a

shows the results of the damping ratio versus

frequency for different Young‘s modulus 85GPa,

95GPa, 105GPa, 115GPa, 125GPa and135GPa.

Figure 3b: Variation of the damping coefficient

with friction coefficient at frequency 12 kHz.

It can be seen that the major squeal frequency does

not change for different values of Young‘s modulus

for the disc. The value of the major squeal

frequency is approximately 12 kHz. As Young‘s

modulus is increased and hence as the stiffness of

the disc is increased, the value of the damping

coefficient decreases. Similar evaluations have

been carried out by Liu et al [11] Figure 4b,

presents the damping coefficient versus Young‘s

modulus of the disc at a frequency of 12 kHz.

Figure 4a: Variation of the damping coefficient

with frequency for different Young’s modulus of

the disc.

It is found that larger disc stiffness can reduce the

squeal propensity of the disc system. This can be

looked upon as increasing the mechanical

impedance of the rotor and therefore making it

more resistive in responding to input forces and

reduce the vibration magnitude; as a result, the

squeal propensity of the disc system can be

reduced.

Figure 4b: Variation of the damping coefficient

with young’s modulus of the disc at

frequency 12 kHz

CONCLUSION

Friction-induced disc brake squeal is investigated

using the Abacus software, which combines a

nonlinear static analysis and a complex Eigen value

extraction method. The nonlinear effects can be

taken into account in the preloading steps in order

to more accurately friction induced damping taken

into account at which a complex Eigen value

analysis is performed. The parametric analysis

shows that significant pad bending vibration may

be responsible for causing the disc brake squeal

and the major squeal frequency is approximately 12

kHz for the existing disc brake system. The effects

of the friction between the pads and the disc, the

stiffness of the disc, and the stiffness of the back

plates of the pads on disc squeal are significant, but

the effects of the hydraulic pressure on disc squeal

are not obvious. Parametric study shows that, if the

Young‘s modulus of the disc is larger, the system is

more stable, and, if the Young‘s modulus of the

back plate of the pads is larger, the system is more

unstable.

99



ACKNOWLEDGEMENTS

The authors would like to thank Dr Abd Rahim

Abu Bakar of Universiti Teknologi Malaysia for

his valuable advice and comments towards the

preparation of this manuscript.

REFERENCES

1. Kung, S.W., Dunlap, K.B. and Ballinger R.S.,

2000. ―Complex Eigen value analysis for

reducing low frequency brake squeal‖ SAE

Technical Report

2. Dunlap, K.B., Riehle, M.A. and Longhouse,

R.E., 1999. ― An investigative overview of

automotive disc brake noise‖ SAE Paper

3. Chen, F., Chern, J. and Swayze, J., 2002.

―Modal coupling and its effect on brake

squeal‖ SAE Paper

4. Mills, H.R., 1938. ―Brake squeal ‖ Technical

Report 9000 B. Institution of Automobile

Engineers

5. Liles, G.D., 1989. ―Analysis of disc brake

squeal using finite element methods ‖ SAE

Technical Paper

6. Bajer, A., Belsky, V. and Zeng, L.J., 2003.

―Combining a nonlinear static analysis and

complex eigenvalue extraction in brake squeal

simulation‖ SAE Paper

7. Lee, L., Xu, K., Malott, B., Matsuzaki, M.

and Lou, G., 2002. ―A systematic approach to

brake squeal simulation using MacNeal

method‖ SAE Paper

8. Blaschke, P., Tan, M. and Wang, A., 2000.

―On the analysis of brake squeal propensity

using finite element method‖ SAE Paper

9. Kung, S.W., Stelzer, G., Belsky, V. and Bajer,

A., 2003. ―Brake squeal analysis

incorporating contact conditions and other

nonlinear effects‖ SAE Paper

10. AbuBakar, A. R. and Ouyang, H., 2006.

―Complex Eigen value analysis and dynamic

transient analysis in predicting disc brake

squeal‖ Int. J Vehicle Noise Vib. 2 (2) 143–

155.

11. Liu, p., Zheng, H., Cai, C., Wang, Y.Y., and

Ang, K., 2007. ―Analysis of disc brake squeal

using the complex Eigen value method‖

Applied Acoustic 68, 603–615.

12. Mario, T. J., Samir N.Y. and Roberto, J.,

2008. ―Analysis of brake squeal noise using

the finite element method: A parametric

study‖ Applied Acoustics 69, 147–162

13. Inman, D.J., 2006. ―Vibration with Control‖

Wiley, New York.

100



N-2010-S06-30

Effect of EGR on Emission Characteristics of Crude Rice

Bran Oil Blend as a CI Engine Fuel at Higher Injection

Pressure

S. Saravanan 1, J. Venkatesan

2, G. Nagarajan

3

Keywords: Alternate fuel, IC engine, Emission characteristics, Blended fuel

ABSTRACT

In the present work emission characteristics of

crude rice bran oil (CRBO) blend was investigated

as a CI engine fuel with EGR at higher fuel

injection pressure. CRBO blend contains 20 % of

CRBO and 80 % of No.2 petroleum diesel on

volume basis. NOx, smoke density, UBHC, CO

and brake thermal efficiency were presented at

various loads for 15 % of EGR at a fuel injection

pressure of 230 bar. It was observed that the NOx

emission was decreased significantly at higher

loads without any increase in smoke and UBHC

emission. It was also observed that the CO

emission was increased and brake thermal

efficiency of the engine was reduced marginally as

a result of the combined effect.

INTRODUCTION

Vegetable oils and their derivatives were

considered as a promising alternate fuel for diesel

as they are renewable in nature and environmental

friendly. However their NOx emission is higher

than diesel which needs further research wok to

find a solution for the same [1-4]. Retardation of

fuel injection timing and EGR are well known

methods to reduce the NOx emission of diesel

engine [5]. However it was reported that, these

methods will increase the smoke emission of the

engine [6-8]. In this work an attempt was made to

reduce the NOx emission of vegetable oil blend

through EGR. To reduce the increase in smoke

density as a result of EGR, fuel injection pressure

was increased [5]. Hence the main objective of the

present investigation is to investigate the combined

effect of EGR and injection pressure in the

emission characteristics of the engine fuelled with

vegetable oil blend.

Vegetable oil used in this investigation is high free

fatty acid (FFA) crude rice bran oil (CRBO) which

is a non-edible vegetable oil derived from rice

bran[9]. CRBO with high FFA content is a non-

edible vegetable oil which can be utilized in CI

engine in blended form with diesel as an alternate

to diesel fuel [10] .High FFA CRBO has

comparable properties as that of diesel [11] and the

properties of CRBO blend compared with diesel

are given in Table 1.

EXPERIMENTAL SETUP

Schematic diagram of the experimental set-up is

shown in Figure 1. The technical specifications of

the engine used in this investigation are given in

Table 2. A piping arrangement was provided to tap

the exhaust gases from the exhaust pipe and to

connect it into the inlet air flow passage.

Table 1. Properties of CRBO blend and Diesel l

Property Testing Method Diesel B3

Kinematic

Viscosity (Cst)

Redwood

Viscometer 3.63 12.34

Calorific Value

(KJ/Kg) ASTM D 240-02 43000 40136

Aniline Point

(Deg C) ASTM D611 74 61

Volatility (%) ASTM D86-00A 90 85

101



Figure 1 Layout of Experimental Setup

EXPERIMENTAL PROGRAMME

Table 2 Specifications of engine

Make Kirloskar

Model TAF 1

Type Direct injection, air

cooled

Bore × Stroke (mm) 87.5 × 110

Compression ratio 17.5:1

Capacity 0.661 lit

Rated power 4.4 kW

Rated speed 1500 rpm

Start of injection 23.4º bTDC

Injector operating

Pressure 200 – 205 bar

% EGR was calculated by using the expression

100%1

21 XM

MMEGR

Where M1= mass of air without EGR

M2= mass of air with EGR

The spring tension of the injector needle with

setting screw was varied to get the different fuel

injection pressure.

TESTING PROCEDURE-Tests were conducted on

the engine to determine the effect of EGR and

injection pressure on the objective. Tests were

carried out at various loads starting from no load to

full load condition at a constant rated speed of 1500

rpm. At each load, the fuel flow rate and the

composition of exhaust gases were recorded under

steady state conditions. Various constituents of

exhaust gases like unburned hydrocarbons

(UBHC), carbon monoxide (CO), nitrogen oxides

(NOx), and carbon di-oxide (CO2) were measured

with a 5-gas MRU 1600 Delta exhaust gas

analyzer. The engine was first operated at normal

operating condition with CRBO blend to generate

the baseline data and then the test was conducted

with EGR and the measurements were made as

before.


This section describes the combined effect of EGR

and injection (CRBO blend-CE) on the

performance and emission characteristics of CRBO

blend by comparing the same with normal

operating condition.

Figure 2 variation of NOx emission

Figure 2 shows the variation of NOx emission with

load as a result of EGR and injection pressure. It

was observed that the NOx emission was reduced

significantly as a result of combined effect. The

reduction in NOx was higher when the load on the

engine was more than 50%. This is due to the

decrease in leanness of the mixture as a result of

EGR. Since CRBO is an oxygenated fuel, it

supplies additional oxygen to the mixture which

makes the mixture lean. This increase in leanness

150

250

350

450

550

650

750

850

950

1050

1150

0 25 50 75 100

NO

x in

pp

m

Load in %

CRBO blend

102



was reduced by recycling exhaust gases since it

reduces the mass of air admitted into the cylinder

and makes the mixture closer to stoichiometric and

hence improved combustion which increase the

NOx emission. At higher loads this leanness was

getting decreased by increased fuel flow with load.

Figure 3 variation of smoke density

Variation of smoke density with load as a result of

combined effect is shown in Figure 3 by comparing

it with normal operating condition. It was observed

that the smoke density was decreased significantly

at all operating conditions except at full load. This

decrease in smoke density was due to the increased

fuel injection pressure which enhances the fuel

atomization and vaporization and hence the

decreased smoke density.

Figure 4 variation of UBHC emission

Variation of UBHC with load as a result of

combined effect is shown in Figure 4 by comparing

it with normal operating condition. It was observed

that the UBHC emission was not increased by the

combined effect of EGR and fuel injection

pressure. This is mainly due to the increase in fuel

injection pressure which reduces the particle size of

the fuel and enhances the atomization of fuel. This

causes every particles to take part in the

combustion process and hence reduction in UBHC.

Figure 5 variation of CO emission

Variation of CO emission with load as a result of

combined effect was shown in Figure 5 by

comparing it with normal operating condition. It

was observed that the CO emission was increased

significantly as a result of the combined effect.

Recycling exhaust gases reduces the oxygen

availability which will lead to incomplete

combustion in some regions of the combustion

chamber and hence CO formation.

Figure 6 variation of brake thermal efficiency

Variation of brake thermal efficiency with load as a

result of combined effect was shown in Figure 6 by

comparing it with normal operating condition. It

was observed that the combined effect reduces the

103



thermal efficiency of the engine marginally. As a

result of exhaust gas recirculation, the maximum

temperature attained in the engine combustion

chamber was reduced. This reduces the availability

of heat energy to be converted into mechanical

work. As a result of this a marginal decrease in

thermal efficiency was observed for CRBO blend.

CONCLUSION

From the experimental results it is concluded that

EGR is an effective way to reduce the NOx

emission of CI engine. To reduce the increase in

smoke density as a result of EGR, fuel injection

pressure can be increased which produces lower

UBHC and smoke emission than normal operating

condition. Brake thermal efficiency was reduced

marginally with this combined effect. Hence

through EGR at an increased fuel injection

pressure, the NOx emission of CRBO can be

controlled without any increase in smoke emission

with minor power loss.

REFERENCES

1. Muzio LJ and. Quartucy GC, Implementing

NOx control: research to application, Prog.

Energy Combust. Sci., 1997; 23: 233-266.

2. Watanabe H, Tahara T, Tamanouchi M, Iida

J, Study of the effects on exhaust emissions in

direct injection diesel engines: Effects of fuel

injection system, distillation properties and

cetane number, JSAE Review 1998 ;19(1) :

21- 26

3. IIcıng€ur Y, Altiparmak D, Effect of fuel

cetane number and injection pressure on a DI

Diesel engine performance and emissions,

Energy Conversion and Management 2003 ;

44(3) :389–397

4. Abd-Alla GH, Using Exhaust Gas

Recirculation In Internal Combustion

Engines: A Review, Energy Conversion And

Management 2002; 43(8):1027-1042

5. Henein NA, Analysis Of Pollutant Formation

And Control And Fuel Economy In Diesel

Engines, Prog. Energy Combust. Sci, 1976; L

(4): 165- 207

6. Bari S, Yu C W, Lim T H , Effect Of Fuel

Injection Timing With Waste Cooking Oil As

A Fuel In A Direct Injection Diesel Engine,

Proceedings Of The Institution Of Mechanical

Engineers; Part D J.Automobile Engg 2004;

218(1):93-104

7. Sayin C And Canakci M, Effects Of Injection

Timing On The Engine Performance And

Exhaust Emissions Of A Dual-Fuel Diesel

Engine. Energy Conversion And Management

2009; 50: 203-213.

8. Saleh He, Effect Of Exhaust Gas

Recirculation On Diesel Engine Nitrogen

Oxide Reduction Operating With Jojoba

Methyl Ester, Renewable Energy

2009;34(10): 2178-2186

9. Saravanan S, Nagarajan G, Lakshmi Narayana

Rao G, Sampath S, Feasibility Study Of

Crude Rice Bran Oil As A Diesel Substitute

In a DI-CI Engine Without Modifications,

Energy for Sustainable Developmen, 2007;

11(3): 83-92


Rao G, Investigation on a non-edible

vegetable oil as a CI engine fuel in sustaining

the energy and environment, Journal of

renewable and sustainable energy 2010; 2,

013108: doi: 10.1063/1.3290178


Rao G, Effect of FFA of Crude Rice Bran Oil

on the Properties of Diesel Blends. J Am Oil

Chem Soc. 2008; 85(8): 663–666

CONTACT

1, 2 Assistant professor, Automobile Engineering,

Sri Venkateswara College Of Engineering,

Sriperumbudur, Chennai

3 Professor, Mechanical Engineering, College of

Engineering, Guindy, Chennai

104



N-2010-E07-33

Control of Regulated Emissions Using Intelligent Cylinder

Deactivation

1 Sanjeev Bhushan R and Nagarajan G

College of Engineering, Anna University Chennai

Keywords: IC engine, emission, engine design,

ABSTRACT

Automotive manufacturers all around the world

have been forced into looking for various ways of

their engines meeting the emission standards.

Emphasis on development of technology has

shifted towards electronic control of the entire

engine operation. One concept that has evolved

over the years is that of Cylinder Deactivation.

Many manufacturers have gone in for deactivation

of a particular number of cylinders of an engine at

low load; low speed operating conditions in order

to meet emission standards. The usage of such a

concept is new in India. The emission

characteristics for operation under various cylinder

combinations with Intelligent Cylinder

Deactivation have been compared and analysed.

INTRODUCTION

Manufacturers around the world have been forced

to look for technologies providing them with the

maximum efficiency for their engines due to the

growing demand for fossil fuels with the ever

growing need to meet emission norms. One such

concept that has evolved over the years has been

that of Cylinder Deactivation. Many manufacturers

have gone in for deactivation of a particular

number of cylinders of an engine at particular

operating conditions in order to derive efficiency

[5]. The usage of such a concept is new in India.

There are very few vehicles plying on the Indian

roads with engine systems incorporating cylinder

deactivation. The major component in flawless

operation of such a system is the Electronic Control

Unit (ECU). The ECU takes into account all the

factors that affect the performance of an engine

before sending out signals to the corresponding

injectors and spark plugs. The ECU processes the

input data available and looks at the various

performance maps, fuel injection timing tables and

spark timing tables for the corresponding engine

speed, and load, thereby achieving a very close

control over the operation of an engine.

The technology available today looks at

deactivation of a particular fixed set of cylinders

under favourable load conditions. This is well-

suited for European conditions where the vehicle

operates with a set of cylinders deactivated only for

a small duration of time. In India, the operation of

the engine under idle conditions is very high which

leads to avoidable emissions from the engine when

waiting at a traffic signal or when parking by the

side of the road. The current work has been carried

out with an intention to reduce unnecessary use of

power when the application‘s demand is far lower

than that is being delivered resulting in reduced

engine emissions under idle conditions.

Hazler & Zwetz have developed a method for

controlling operation of an multi-cylinder engine

by monitoring a parameter that is associated with

the engine operation. The fluctuation of the

selected parameter has been monitored and if the

fluctuation exceeds the threshold, at least one of the

cylinders is deactivated. The experiments have

been carried out on a C.I. Engine.[1]

Armin Herold and Peter Lückert have developed a

set-up of a super-chargeable internal combustion

engine with cylinder cut-off, comprising of two

cylinder groups of which the first group operates

over the entire operating range and a second group

which is cut-off or connected as per the demand.

An exhaust-gas turbocharger is arranged

exclusively for supercharging the first group which

has the first charging-air feed and exhaust-gas

105



discharge. A mechanical supercharger is arranged

for supercharging the second group. It can also be

coupled to the drive train by a switchable clutch. At

transition from part-load operation to full-load

operation, the cylinders of the second cylinder

group initially require a start-up phase, during

which the associated exhaust-gas turbocharger also

begins operation.[2]

Yutaka Ide has devised a control system for

cylinder cut-off and which at all times, maintain

favourable emission characteristics upon switching

from the partial cylinder operation to a full cylinder

operation. The system also maintains a satisfactory

fuel consumption rate by conducting the partial

cylinder operation to the utmost. Cylinders on the

right bank are selected for cut-off. Exhaust gases

from both banks are purified by two catalyst units

separately. The controller estimates the catalyst

temperature on the right bank and enables the

cylinder deactivation mechanism when the

temperature is above a pre-determined

temperature.[4]

During part-load operation, Mercedes have

developed a system that deactivates four of the

eight cylinders, and brings them back into

operation when greater performance is required.

When the system shuts off four of the eight

cylinders, fuel consumption under the New

European Driving Cycle (EUDC) is reduced by

about 7% which can be further increased under

other driving conditions.[5]

On Indian roads, engines are operated in the low

speed; low load ranges for a prolonged period

often. Under this prolonged usage of a fixed set of

cylinders, the wear and tear of the piston and

cylinders which are selected to be deactivated and

the other cylinders which are running all the time

when the engine is running will be uneven. This

leads to the replacement of the cylinder parts to be

scattered and uneven. With a suitable mechanism to

make the wear and tear in the engine components

even, the problem can be solved.

The current paper focuses on reduction of regulated

emissions whilst taking into account the Indian

driving conditions. The work has been carried out

on a Hyundai Santro Epsilon engine.

METHODOLOGY

By invoking a mechanism to deactivate the

cylinders as various groups, the otherwise

imminent uneven wear and tear could be avoided.

The mechanism is invoked by the use of a

secondary interface. The secondary interface is a

micro-controller which controls the signals flowing

from the ECU to the Fuel Injectors. The main issue

that has to be tackled is that of vibration that could

be creeping in the system due to the cylinders being

deactivated.

The objective being reduction without any change

in the existing mechanism, the cylinders are

deactivated. To avoid the problem of vibration due

to the deactivation, the inner cylinders are grouped

together and the outer cylinders are grouped

together. The required cylinder set is activated by

controlling the flow of signals from the ECU to the

fuel injectors by usage of a secondary interface

between the ECU and the fuel injectors. The

deactivation system is operational only under low

speed conditions (< 2000 rpm) when there is no

external load on the engine.

EXPERIMENTAL SET-UP

A multi-cylinder, four-stroke, electronic fuel

injection, liquid cooled petrol, Hyundai Epsilon

engine, was used to carry out the tests. The

specifications of this engine are given in Table 1.

The tests were conducted under no-load conditions.

Figure 1 Schematic of Engine Set-up

106



Table 1 Engine Specifications

Engine Make Hyundai Epsilon (G4HC)

Configuration In-line

No. of Cylinders 4

No. of Valves 12

Valve Actuation Single Overhead Camshaft

Fuel System Multi-point Electronic Fuel

Injection (EFI)

Bore x Stroke 66 mm x 73 mm

Displacement 999 cc

Power Output 44.87 kW @ 5700 rpm

Torque 110 N-m @ 3000 rpm

The ECU is the heart of the system controlling

almost all the parameters by getting inputs from the

TPS, CPS, MAP, IAT, and CWT Sensors. The

spark timing and injection duration and timing are

varied according to the load and speed of the

engine through a electronically controlled spark

advance mechanism. The CWT Sensor indicates

the temperature of the engine coolant which

decides on the flow rate of the coolant sent to the

engine cooling water jacket. The secondary

interface (Micro-Controller) is connected between

the ECU and the Fuel Injector connections. This

enables the control over the signals being sent to

the respective injectors thereby enabling Intelligent

Cylinder Deactivation (ICD). The existing engine

set-up is as shown in Figure 1.


The experiments have been conducted under cold-

start as well as warmed up state for no load

conditions. Tests were conducted for normal

operation as well as various active cylinder

combinations to determine the most feasible and

profitable combination.

HYDROCARBONS (HC) - Under normal

operating conditions, the HC emissions are around

300 ppm at 1200 rpm and follow an increasing

trend with speed. When moving over the ICD

operation, the HC emissions reduce to a third of

that under normal operation. The engine speed

drops by 400 rpm under similar throttle opening

conditions. The reduced engine speed leads to an

improved volumetric efficiency resulting in more

quantity of air available for combustion of the fuel

that is being injected in the port along with longer

time availability for combustion. As shown in

Figure 2, a reduction of 60 % to 67 % in HC

emissions under all throttle openings is observed

with the implementation of ICD methodology.

Figure 2 Variations in Hydro-Carbon Emissions

CARBON MONOXIDE (CO) - Carbon monoxide

is generated in an engine when the engine is

operated with a fuel-rich equivalence ratio. When

there is insufficiency of oxygen to convert all

carbon to CO2 , some fuel does not get burned

completely and some carbon ends up as CO.

Amount of CO in the exhaust is an indication of

lost chemical energy as it can be combusted to

supply additional thermal energy by complete

conversion into CO2 [Ganesan V, 2007]. Since the

experiments are done within 4 minutes of starting

of the engine, the engine is running under a rich

fuel-air mixture which leads to CO emissions. With

the implementation of ICD in the engine operation,

the CO emissions are cut down drastically by

around 84 % under idling (closed throttle position)

and by a maximum of 92 % as shown in Figure 3.

Figure 3 Variations in CO emissions

0

100

200

300

400

500

600

700

800 1000 1200 1400 1600 1800

HC

, p

pm

Speed, rpm

Normal

ICD

0

0.5

1

1.5

2

2.5

3

3.5

800 1000 1200 1400 1600 1800

CO

, %

Speed, rpm

Normal

ICD

107



CARBON DI-OXIDE (CO2) - Carbon dioxide has

been an influencing factor in global warming

problem arising at every possible regulatory

meeting. The CO2 emissions should therefore be

under control. Although they are an indicator of

good combustion, the quantity of CO2 that is being

emitted from the tail-pipe must be kept under

control under all conditions. Carbon dioxide

emissions have become an area of concern ever

since the global warming scenario came into light.

The CO2 emissions from the engine under normal

operating conditions ranges between 8.9 % and

12.8 % within the speed range of 1200 rpm to 1700

rpm. The figure reduced by 83 % with the

implementation of Intelligent Cylinder

Deactivation with a drop of 400 rpm. With increase

in throttle opening the improvement reduced to 51

% at 1350 rpm under ICD operation as shown in

Figure 4.

Figure 4 Variation in CO2 emissions

This could be attributed to more oxygen

availability for combustion of the increased

quantum of fuel that is being supplied to the engine

cylinder via the port fuel injectors.

Figure 5. Variation in Emission of Oxides of

Nitrogen

EXPERIMENTAL SETUP

Schematic diagram of the experimental set-up is

shown in Figure 1. The technical specifications of

the engine used in this investigation are given in

Table 2. A piping arrangement was provided to tap

the exhaust gases from the exhaust pipe and to

connect it into the inlet air flow passage.

OXIDES OF NITROGEN (NOX) - Regulation of

emission of Oxides of Nitrogen has become

stringent by the year and modern engines make use

of fast-burn combustion chambers which reduce the

concentration of NOx in the exhaust with reduced

combustion time [Ganesan V, 2007]. Normal

engine operation leads to 200 ppm NOx in the

exhaust under closed throttle position. Figure 5

shows that this value reduces by 50 ppm under ICD

operation but the speed reduces by 400 rpm. The

increase in NOx can be related to the better

combustion caused by more oxygen availability

and longer combustion duration which is indicated

by the higher CO2 formation which leads to higher

in-cylinder temperatures under ICD operation.

CONCLUSION

The tail-pipe emissions of a multi-cylinder

automotive S.I. Engine under normal

operation, normal cylinder deactivation and

intelligent cylinder deactivation operation are

investigated and the salient features of the

investigation are summarised below:

The switching period between the various

combinations could be reduced in order to

eliminate the surge in engine speed as well as

engine emissions.

Under similar throttle opening conditions,

with a drop of 400 rpm, HC, CO, CO2, NOx

emissions are reduced by a maximum of 50

%, 80 %, 67 %, and 30 % respectively when

operating with Intelligent Cylinder

Deactivation as compared to normal engine

operation.

From the observations made in the current

work, Intelligent Cylinder Deactivation is

found to be effective in reduction of regulated

emissions as compared to conventional

operation. Hence, it can be concluded that

Intelligent Cylinder Deactivation is an useful

addition to the engine system. ICD can be

0

5

10

15

800 1000 1200 1400 1600 1800

CO

2, %

Speed, rpm

Normal

ICD

0

200

400

600

800

800 1000 1200 1400 1600 1800

NO

x, p

pm

Speed, rpm

Normal

ICD

108



used in existing as well as new engine systems

as a means to reduce regulated emissions in

order to meet stringent emission norms.

ACKNOWLEDGMENTS

The author would like to thank the following

individuals for their assistance in putting this paper

together: Satish Kumar R G, and Chandrasekar R. I

would also like to thank my co-author,

Dr.G.Nagarajan, without whose support and inputs

this study wouldn‘t have been possible.

REFERENCES

1. Hasler, G. S., Zwetz, D. L., "Cylinder Cut-out

Strategy for Engine Stability," United States

Patent 7073488, July 11, 2006.

2. Herold, A., Lückert, P., Super-chargeable

Internal Combustion Engine with Cylinder

Cut-off, United States Patent 6158218, 12th

December 2000.

3. Ishiyama, M., Nishida, K., Okada, T., Sen, N.,

Sugiyama, A., Tomokuni, Y., Yamashita, K.,

"Control System for Cylinder Cut-off Internal

Combustion Engine," United States Patent

7308962, December 18, 2007.

4. Ide, Y., Controller for Cylinder Cut-off Type

Internal Combustion Engine, United States

Patent 6408618, June 25, 2002.

5. Automotive Engineering International Online,

"Mercedes-Benz launches cylinder cutout,"

6. http://www.sae.org/automag/newenginereview/

mercedes.htm

DEFINITIONS, ACRONYMS,

ABBREVIATIONS

ECU : Electronic Control Unit

ICD : Intelligent Cylinder Deactivation

TPS : Throttle Position Sensor

CPS : Crank Position Sensor

MAP : Manifold Absolute Pressure

IAT : Inlet Air Temperature

CWT : Coolant Water Temperature

ECT : Engine Coolant Temperature

TFC : Total Fuel Consumption

HC : Hydrocarbons

CO : Carbon Monoxide

CO2 : Carbon Di-oxide

NOx : Oxides of Nitrogen

CONTACT

1E-mail: [email protected]

109



THE GREEN CAMPUS

110



111



112



113

National Conference - NCAAT 2010

Documents

Transcript of National Conference - NCAAT 2010