research on vehicle technologies_u08

First Edition 2008 © SRITHAR RAJOO 2008

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical including photocopy, recording, or any information storage and retrieval system, without permission in writing from Universiti Teknologi Malaysia, Skudai, 81310 Johor Darul Tak'zim, Malaysia. Perpustakaan Negara Malaysia Cataloguing-in-Publication Data Research on vehicle technologies 2008 / editor srithar Rajoo ISBN 978-983-52-0550-7 1. Automobile--Design and consruction. 2. Automobiles--Technological innovations. I. Srithar Rajoo, 1975-. 629.23

Pereka Kulit: MOHD. NAZIR MD. BASRI

Diatur huruf oleh / Typeset by SRITHAR RAJOO & RAKAN-RAKAN

Fakulti Kejuruteraan Mekanikal Universiti Teknologi Malaysia

81310 Skudai Johor Darul Ta'zim, MALAYSIA

Diterbitkan di Malaysia oleh / Published in Malaysia by

PENERBIT UNIVERSITI TEKNOLOGI MALAYSIA

34 – 38, Jalan Kebudayaan 1, Taman Universiti, 81300 Skudai,

Johor Darul Ta'zim, MALAYSIA. (PENERBIT UTM anggota PERSATUAN PENERBIT BUKU MALAYSIA/

MALAYSIAN BOOK PUBLISHERS ASSOCIATION dengan no. keahlian 9101)

Dicetak di Malaysia oleh / Printed in Malaysia by UNIVISION PRESS

Lot 47 & 48, Jalan SR 1/9, Seksyen 9 Jln. Serdang Raya, Tmn Serdang Raya

43300 Seri Kembangan, Selangor Darul Ehsan MALAYSIA

Contents Automotive Turbocharging Srithar Rajoo and Ricardo Martinez-Botas PD Control Performance for Ratio Control of an EMDAP- CVT Bambang Supriyo, Kamarul Baharin Tawi, Hishamuddin Jamaluddin and Sugeng Ariyono Brake Squeal Control Using An Active Technique Musa Mailah, S.M Hashemi-Dehkordi and A.R. Abu Bakar Reduction of Emissions in DI Diesel Engine Using High Turbulence Combustion with EGR And Pilot Injection Wira Jazair, Yoshiyuki Kidoguchi and Kei Miwa Performance and Emission Evaluations of a Prototype Stepped-Piston Engine Using Carburetor and Direct Fuel-Injection Systems Azhar Abdul Aziz, Zulkarnain Abdul Latiff, Mohd Fawzi Mohd Ali, Mohd Farid Mohammad Said and Mazlan Said The Development of Gray Cast Iron Cylinder Block for Passenger Vehicle Raja Mazuir Raja Ahsan Shah Identification of Vehicle Suspension System Using Particle Swarm Optimization with Neural Network Musa Mailah, Gigih Priyandoko and Hishamuddin Jamaluddin

1 23 41 59 89 105 121

240 Advances in Manufacturing and Industrial Engineering (2008)

PREFACE

This book puts together some selected research works on vehicle technologies conducted by the staffs of Universiti Teknologi Malaysia (UTM), particularly the members of the Faculty of Mechanical Engineering (FME). UTM is the first institution in Malaysia to offer a bachelors level Automotive Engineering programme. FME has a number of automotive experts who are active in their field of research with continuous emphasis on disseminating results and findings. These experts are mainly belong to the Department of Automotive Engineering (DAE), an entity within the faculty. The research works of DAE are grouped into 3 main areas, namely Powertrain Engineering, Vehicle Dynamics & Structures and Electronics & Control. The chapters in the book in general represent the research focus area of the department.

The first chapter presents an overview of the development in turbocharging technology. Turbocharger has been proposed recently as one of the key element in engine development as the current demand in automotive industries moves toward emission reduction and downsizing. The chapter provides the readers with the insight of the working principles and different types of turbochargers. Furthermore, few significant researches in the field are discussed with elements of the author’s own work.

The second chapter presents the development of a Continuous Variable Transmissions (CVT), focusing the discussion on the control strategy. The presented CVT work is fully conducted in

vi Preface

UTM with the authors successfully granted few patents for their invention.

The third paper is also on the control aspect of a vehicle, where the authors presents their work on brake squeal. A novel Active Force Control based scheme has been developed to effectively suppress the vibration and noise (squeal) emanating from a disc brake system.

The fourth chapter presents the work on diesel technology. The authors described their method to reduce emission in a Direct Injection (DI) diesel engine. They have utilized high turbulence with Exhaust Gas Recirculation (EGR) and pilot injection to achieve the target emission reduction. This is very relevant to the inevitable concern of clean environment and making automotive as least pollutant as possible.

The fifth chapter also elaborates on the emission of an engine. However, here the authors presented their work on a small 2-stroke gasoline engine. The engine was fully developed in UTM which includes a direct fuel injection system to improve performance and reduce emission.

The sixth chapter presents the industrial view to an automotive engineering problem. The author describes the development of gray cast iron cylinder block for passenger cars. Two major concerns in the development have been studied and solutions proposed to resolve them.

The last chapter in the book describes the use of intelligent control method to optimize an automotive system. The authors used particle swarm optimization with neural network to model a vehicle suspension system. This has been shown as an alternative to the conventional way of modeling passive vehicle suspension systems.

All of these chapters are presented as to share the knowledge and research findings with the readers. The editor hopes that this book serves its purpose in disseminating knowledge and creating enthusiasm among the readers in the area of Automotive Engineering. The editor would like to convey his appreciation to all

vii Preface

the authors for their contributions and to Dr. Rahim Abu Bakar for editorial assistance. THANK YOU.

Srithar Rajoo Department of Automotive Engineering Faculty of Mechanical Engineering Universiti Teknologi Malaysia 2008

1

AUTOMOTIVE TURBOCHARGING Srithar Rajoo

Ricardo Martinez-Botas

1.1 INTRODUCTION The future trends of automotive engine are universally toward down-sizing, higher power density and above all lower carbon emissions. Among many technologies revolutionizing automotive development, turbocharging is considered as a significant enabler to meet the ever increasing future demands. Uchida (2006) provided a good discussion on the future trends for the automotive industry and the inherent role of turbocharging, with focus on the Toyota research developments. Figure 1.1 shows the demand for specific power to increase to 70 kW/l and CO2 emission to reduce to 115 kg/km by the year 2010. Achieving the goal, according to Uchida (2006), will need technological steps forward with turbocharging enhancement as the main player. These views are also shared by Shahed (2005) in his article discussing the general demand and importance of turbocharging for the current and future automotive powertrain. Down-sizing and emission reduction were the main driving force behind the significant development of turbodiesels in Europe and similar development are predicted for the United States automotive industry.

Shahed (2005) suggested the development of Variable Geometry Turbocharger (VGT) has contributed positively to the evolution of diesel engines. This includes the capability of VGT to boost over wider speed range and its compliment to the Exhaust Gas Recirculation (EGR) operation in reducing harmful emission. Apart from VGT, several methods of turbocharging are proposed

2 Research on Vehicle Technologies (2008)

and tested currently by manufacturers, which includes the motor assisted VGT and two-stage turbocharging (Shahed, 2005 and Uchida, 2006). In a motor assisted VGT, an electric motor is used to regulate the nozzle position to provide almost an instant boost when necessary. Meanwhile, a two-stage turbocharging provides a wider operating range with higher pressure ratio and air-flow. However, VGT remains the well proven technology readily available and in many cases considered indispensable in the effort to meet future engine demands.

Shahed (2005) and Uchida (2006) suggested very promising potential of turbocharging in the evolution of modern automotive powertrain. These include the development of down-sized gasoline engines, engine-motor hybrid vehicles and fuel cells. Turbochargers are expected to continually contribute to the universal effort of energy saving and global warming reduction. 1.2 HISTORY OF TURBOCHARGING The concepts of turbo charging dates back to the time of internal combustion engine development, when Gottlieb Daimler and Rudolf Diesel investigated the possibility of increasing the power output and reducing the fuel consumption of their engines by pre-compressing the combustion air. In 1925, the Swiss engineer Dr. Alfred Büchi was the first to be successful with exhaust gas

Figure 1.1 Automotive powertrain power and emission demands for the future (extract from Uchida, 2006)

Automotive Turbocharging 3

turbocharging and achieved engine power improvement. Figure 1.2 shows the cross sectional cut of the turbocharging unit built by Dr. Alfred Büchi.

Turbocharging begins with truck engines in the automotive industry, when the first turbocharged engine for trucks built by the Swiss Machine Works Saurer in 1938. As for passenger cars, the Chevrolet Corvair Monza and the Oldsmobile Jetfire were the first to be turbocharged in 1962/63, but did not last long due to reliability problems. Turbocharged diesel engines became popular after the oil crisis in 1973, and the consequent stringent emission regulations from the 80s saw the significant development till today almost all heavy duty diesel engines are turbocharged. Traditionally turbocharging has been associated with power improvement, but suffered from problems such as reliability and turbo-lag. However, continual development and introduction of new techniques such as variable geometry turbocharging has resolved most of the classical problems. In the current environment, the primary reason for turbocharging is to reduce fuel consumption and emissions.

As for the market domination only about 6% of the petrol cars are currently turbocharged, while almost 100% of diesel cars run on turbocharger (Beecham, 2003). While the lack of application in petrol engine prompts future potential, the

Figure 1.2 First turbocharger designed by Dr. Alfred Büchi


domination in diesel engine does not in any measure means turbocharging has reach its peak. Relative to the long history of turbochargers, its technology and improvement is still away from a saturated stage. Manufacturers and researchers are continuously seeking new methods of improvement to further enhance the turbocharger application.

The discussion of the turbocharging history is largely based on information gathered from BorgWarner Turbo & Emission Systems (History, 2002-2007). 1.3 WORKING PRINCIPLES OF TURBOCHARGERS Internal combustion engine is essentially an air-breathing machine, inducing air for combustion and consequently exhaust the by-product in a well tuned continuous cycle which produces useable power. A turbocharger enhances the breathing process by feeding high density air into an engine for better combustion process. This enables a turbocharged engine to produce more power and torque compared to an equivalent natural aspirated engine. A typical turbocharger consists of a turbine and a compressor coupled to a common shaft (Heywood, 1988) as shown in Figure 1.3. The working principle of a turbocharger is the turbine extracts exhaust gas energy to power the compressor which in return increases the density of the charged-air delivered to the engine cylinders. The general effectiveness of a turbocharger is in maintaining a positive pressure gradient across an engine intake-to-exhaust. In a fixed geometry turbocharger the turbine geometry is fixed thus limiting the maximum exhaust flowing through it before the back-pressure increases to create negative pressure gradient.

As can be seen in Figure 1.3, a wastegate is used to bypass the excess exhaust gas in aim to consistently maintain positive pressure-gradient, but with the penalty of losing usable energy. On the other extreme, very low exhaust flow will be inadequate for the turbine to spin and deliver the power required to meet a driver’s demand. One classic disadvantage of conventional fixed geometry


turbocharger is the finite time required for the turbine to spin to the required power level, which is referred as turbo-lag. This is more severe as the size of the turbocharger increases. Thus a turbocharger usually matched to an engine depending on the operating envelope required. A small turbocharger provides good low engine speed boost but mechanically limited at higher speed range. On the other hand, larger turbochargers are good for high speed operation but poor for low engine speed boost due to its inertia. The classic engine-turbocharger matching difficulty is one of the crucial problems faced by the engine and turbocharger manufacturers in the process of achieving higher engine power with least turbo-lag. Variable geometry turbochargers (VGT) has in many ways resolved this problem and used widely especially for diesel engine application. 1.4 VARIABLE GEOMETRY TURBOCHARGER (VGT) Although the concept of Variable Geometry Turbocharging (VGT) is not entirely new (the first examples appearing in the early

Figure 1.3 Schematic of an engine turbocharging system


1960s) their development only gained impetus in the last 25 years. Up until the early 1980s, VGTs were rarely used except on gas turbine plants and experimental turbochargers. The common problems encountered with VGT were reliability (for long periods of time while exposed to high temperature and corrosive exhaust gases), complexity because of the VGT actuation mechanism and control system, and the subsequent high cost (Watson and Janota, 1982). However, recent research has tended to provide acceptable solutions to most of these problems and today VGT has already had a significant impact in the design of diesel engines.

VGT is considered as an efficient approach to the engine-turbocharger matching. A VGT can match for wider engine speed range by varying the turbine inlet geometry, which in essence similar to having finite range of turbine size in one unit. At lower engine speed and load, the turbine inlet area is reduced (similar to a small turbine) to increase momentum impact on the rotor, while at high engine speed the turbine inlet area is increased to avoid over boosting of the turbocharger. With this flexibility, the exhaust gas can be fully utilized for energy extraction in the turbine, otherwise wasted through a waste-gate as in conventional turbochargers. Figure 1.4 shows the improvement in an engine torque with the use of VGT especially at low speed.

Two of the most widely used variable geometry methods in a turbocharger are pivoting nozzle vane (Figure 1.5) and sliding nozzle ring (Figure 1.6) mechanisms. In the pivoting nozzle vane method shown in Figure 1.5, the nozzle ring is connected to a unison ring and pivoted to regulate the nozzle area as well as the vane angle. Meanwhile in the sliding nozzle ring mechanism as shown in Figure 1.6, the vane angle is fixed and the axial movement of the sliding wall regulates the nozzle area. Both the mechanisms are governed by control strategy as an integral component of modern engines, which largely requires multi-parameters coding to achieve optimum compromise between fast torque response, fuel economy, low emissions and engine safety (Moody, 1986).


Figure 1.4 Engine torque improvement with variable geometry turbochargers(VGT) (extract from Uchida, 2006)

Figure 1.5 Pivoting Vane VGT: The vanes are almost closed at low exhaust flow and vice-versa at high exhaust flow

(extract from Bell, 1997)


1.5 PAST RESEARCHES ON VGT Most research work with variable geometry turbocharger involves an engine performance investigation. These include testing with variation of turbine variable geometry techniques and consequent analysis of its effect on the turbo-lag, torque response, fuel consumption, emission level, drivability and the overall performance. Few of the published work on aerodynamic investigation of a variable geometry turbine are by Capobianco and Gambarotta (1992) and Baets et al. (1998).

Watson and Janota (1982) described the benefits of VGT as most beneficial for engines with substantial load and speed changes, which in most cases are road vehicles. However, they stated the high cost and reliability problem of VGT system during the time as the major factors to be resolved before its wide-spread use.

Figure 1.6 Sliding Nozzle Ring VGT: The nozzle ring is least exposed at low exhaust flow, thus reducing the effective area and vice versa at high exhaust flow (courtesy of HOLSET Engineering)


Figure 1.7 shows two research level VGT forwarded by Watson and Janota (1982) to demonstrate its benefits over a fixed geometry unit in terms of improved brake power, specific fuel consumption and reduced emission. This is the time where VGT was still restricted to expensive research level prototype and to some extent niche industrial penetration.

Chapple et al. (1980) conducted an aerodynamic investigation into designing series of radial turbine volutes. These include a ‘variable geometry’ nozzleless casing, where a moving wall was used to create fluid passage with consistent energy and momentum conversion for a range of mass flow rates. The movable wall fluid passage will then deliver the flow into a nozzleless space upstream of the rotor. Chapple et al. (1980) tested only two settings of the variable geometry casing, open and closed wall settings as shown in Figure 1.8. Only partial efficiency

Kunberger (1980)

Figure 1.7 Engine performance improvements with variable geometry turbocharger demonstrated by different researchers as forwarded by Watson and Janota (1982)

Berenyi and Raffa (1979)


improvements were recorded with the variable geometry casing, which was at the low mass flow rate region. However, in overall the work by Chapple et al. (1980) was mainly focused on establishing a systematic design approach for a radial turbine casing.

Flaxington and Szczupak (1982) examined few methods of variable area devices for turbines. These include in general the variable area via regulation of the rotor exit area, volute exit area and volute tongue area. They further investigated the latter two methods, which are shown in Figure 1.9 with its relevant engine performance results. They concluded in general an improved torque, wider speed range and improved transient performance. Flaxington and Szczupak (1982) also discussed the different control strategy for the VG turbine, namely boost control, engine

(a) Open wall configuration (b) Closed wall configuration

Efficiency lower than fixed geometry casing

Better efficiency than fixed geometry casing at

low mass flow rate

Figure 1.8 Nozzleless ‘variable geometry’ casings with open (a) and close (b) wall settings (extract Chapple et al., 1980)


speed control and optimum full/part load control. However, they suggested that no one device or area control method is in overall superior for all application, instead each exhibits benefits in certain operating envelope of the engine. Walsham (1990) presented similar investigation comparing the relative benefits of waste-gated turbine, variable geometry turbine and turbo compounding.

Volute exit area control methods and performance

Tongue area control methods and performance

Figure 1.9 Two general approach of variable area control in a turbine and the variety method within each, with relevant engine performance (extract from Flaxington and Szczupak, 1982)


Walsham (1990) suggested technically the variable geometry turbines are better for transient response as shown in Figure 1.10, as well as the transient emission. Nevertheless, Walsham (1990) demonstrated significant efficiency drop in the VG turbine at closed positions. In addition, the complexity of the VG system is another factor need to be weight upon for an engine application.

Wallace et al. (1986) presented a series of engine testing with a variable geometry turbocharger to establish its benefits. In overall they demonstrated improvement in the engine transient response but with efficient penalty for the turbine. Wallace et al. (1986) documented many related problems faced with the use of variable geometry turbochargers and suggested the need for optimal control strategy to enhance positive engine improvement. Wallace et al. (1986) suggested then the variable geometry turbine is the next stage in turbocharging technology. Hawley et al. (1999) conducted similar investigation as Wallace et al. (1986) but during

Figure 1.10 Engine transient responses with alternative turbocharging techniques (extract from Walsham, 1990)


the phase where VGT has become a standard component in almost all diesel engine applications. Hawley et al. (1999) focused on the control strategy of a VGT for optimal emission without performance penalty. A turbocharger with pivoting nozzle vane turbine was equipped to a direct injection diesel engine. Through manual control of the VGT positions, improvement of up to 45% was achieved in the NOx emission and 10% in the limiting torque, compared to a standard fixed geometry turbocharger, shown in Figure 1.12. A similar engine torque curve at limiting torque condition from Wallace et al. (1986) is shown in Figure 1.11. Comparison of these two investigations provides reader with the extent of development in the variable geometry turbocharging. VGT problems such as reliability and optimal control requirement faced in the 80s are significantly resolved by the end of 90s.

O – Z restricting nozzle throat area

Figure 1.11 Engine torque under limiting torque condition (extract from Wallace et al., 1986)


Capobianco and Gambarotta (1992) presented one of the very few aerodynamic investigations on variable geometry turbines. A variable area turbine and a variable nozzle turbine were tested in steady and pulsating flow conditions to establish its relative aerodynamic performance. The general characteristics of variable geometry turbines were presented such as wider operating range and peak efficiency drop. Nevertheless, the variable geometry turbine efficiency was concluded to be better than the equivalent fixed geometry for the overall operating range, as

Figure 1.12 Engine torque and NOx emission under limiting torque condition (extracted from Hawley et al., 1999)


shown in Figure 1.13. They further suggested the variable nozzle unit to be marginally better than the variable area unit.

Capobianco and Gambarotta (1992) consequently suggested method to correlate the unsteady performance parameters and the equivalent quasi-steady values. However, only the inlet static pressure was measured instantaneously at the inlet and exit of the turbine, while the rest of the parameters were deduced as time-averaged values. A factor taking in ratio between the unsteady time-averaged parameter and the equivalent steady parameter was proposed. However, no conclusive relation could be established based on the factors for the whole testing range presented except for the mass flow rate. A general decaying mass flow rate factor was presented with increasing pulse amplitude.

Figure 1.13 Optimum efficiency of the variable nozzle (VNT), variable area (VAT) and fixed geometry (TB) turbines

(extract from Capobianco and Gambarotta, 1992)


1.6 TYPES OF VGT Many different methods have been proposed over the years since the inception of the variable geometry principles for a turbocharger turbine. However only two such methods have been proven most reliable and since used widely on most current VG turbochargers; these are the variable area through moving wall/sliding nozzle and the variable nozzle through pivoting methods. However, there are few new methods proposed recently as an improvement to the existing variable geometry techniques. Kawaguchi et al. (1999) presented a Variable Flow Turbocharger (VFT), as a novel technique to vary a turbine inlet area, as shown in Figure 1.14. It consists of an additional scroll outside the normal fixed nozzled scroll. The fixed nozzle vane ring is placed at the intermediate region between both the scrolls and a control valve is used to control the air-flow into the outer scroll. The working principles are shown in Figure 1.14 for low and high flow rate. Kawaguchi et al. (1999) suggested the major positive point of the VFT is the single moving component (the control valve) compared to the multiple nozzle vanes movement in a VGT. Kawaguchi et al. (1999) documented a 10 kPa boost pressure improvement in the VFT compared to an equivalent VGT under similar inlet exhaust pressure. They further conducted an engine coupled test and concluded better boost pressure controllability with the VFT compared to a VGT. In more recent years, Pesiridis and Martinez-Botas (2007) presented a new technique of actively regulating a turbine inlet area to adapt to the instantaneous exhaust gas pulsation; called Active Control Turbocharger (A.C.T.), shown in Figure 1.15. This is suggested as an advance technique where a variable geometry turbine operation extended to consider the pulsating nature of the inlet exhaust flow. Marginal power improvement was documented but with efficiency penalty contributed largely by the aerodynamically poor sliding nozzle employed. The improved version of the A.C.T. was presented by Rajoo and Martinez-Botas (2007) with a pivoting nozzle ring, as shown in Figure 1.16.


Many recent investigations on variable geometry

turbocharger involve the control strategy for power or emission optimization. These are in most cases deals with the transient operation of an engine. Stefanopoulou et al. (2000) have analyzed the control aspect of the joint effect of VGT and EGR (Exhaust Gas Recirculation) on diesel engine emissions. A multivariable

(a) Low flow rate

(b) High flow rate

Figure 1.14 Alternative variable geometry turbine method : Variable flow turbine (VFT) mechanism (Kawaguchi et al., 1999)


feedback controller was proposed for emission improvement. Similar studies were done by Nieuwstadt et al. (2000) using different control strategies and evaluating emission quality based on new European drive cycle. Meanwhile Moody (1986) presented the electronic control system of VGT. Brace et al. (1999) and Filipi et al. (2001) documented investigation into the transient response and improvement of an engine with VGT. Brace et al. (1999) reported the experimental response of two alternative designs of VGT system to step pedal inputs. Meanwhile Filipi et al. (2001) developed a new methodology to study the VGT transient response with simulation.

Figure 1.15 Concept of Active Control Turbocharger (A.C.T.); nozzle gap progressively opens towards the peak pressure and closes lower end of the pulse (extract from Pesiridis and Martinez-Botas, 2007)


1.7 CONCLUSIONS This chapter presents the overview of the turbo charging technology and the relevant researches in the past decades. Turbochargers are more than 100 years old, yet its development is still growing with continuous improvement to meet the demand of current automotive and energy market. Many of the problems with turbochargers such as turbo lags have been resolved through

Vanes Driving Ring

Connect to the Shaker

10°

40°

Pin-Armed Lever for pivoting

Figure 1.16 Turbine with the pivoting mechanism of the nozzle vanes (extract from Rajoo and Martinez-Botas, 2007)


technology development as in Variable Geometry Turbochargers (VGT). Turbocharger has been a standard component in almost all diesel engines, where it currently finds increasing consumer market capitalization. Furthermore, as one of the enabling technology to achieve the lower emission and downsizing target, turbochargers are now increasingly utilized for gasoline engines. REFERENCES Baets, J., Bernard, O., Gamp, T., And Zehnder, M., 1998. Design

and Performance of ABB Turbocharger TPS57 with Variable Turbine Geometry. In: 6th International Conference on Turbocharging and Turbochargers, Institution of Mechanical Engineers, London, paper C554/017/98, pp. 315-325.

Beecham, M., 2003. The global market for automotive turbochargers for passenger cars and commercial vehicles. Research Report: ABOUT Automotive, ABOUT publishing limited.

Brace, C.J., Cox, A., Hawley, J.G., Vaughan, N.D., Wallace, F.J., Horrocks, R.W., Bird, G.L., 1999. Transient Investigation of Two Variable Geometry Turbochargers for Passenger Vehicle Diesel Engines. Society of Automotive Engineers (SAE), paper 1999-01-1241.

Capobianco, M., And Gambarotta, A., 1992. Variable geometry and waste-gated automotive turbocharges: Measurements and comparison of turbine performance. Journal of Engineering for Gas Turbines and Power, Vol. 114, pp. 553-560.

Chapple, P., Flynn, P. F. & Mulloy, J., 1980 Aerodynamical design of fixed and variable geometry nozzleless turbine casings. Transactions of the ASME Journal Eng for Power 102, 141–147.

Filipi, Z., Wang, Y., And Assanis, D., 2001. Effect of Variable Geometry Turbine(VGT) on Diesel Engine and Vehicle


System Transient Response. Society of Automotive Engineers (SAE), paper 2001-01-1247.

Flaxington, D. And Szuzupak, D. T., 1982. Variable Area Radial-Inflow Turbine. In: 2nd International Conference on Turbocharging and Turbochargers, Institution of Mechanical Engineers, London, paper C36/82.

Hawley, J. G., Wallace, F. J., Cox, A., Horrocks, R. W., And Bird, G. L., 1999. Variable Geometry Turbocharging for Lower Emissions and Improved Torque Characteristics. Proceedings of the IMechE, Journal of Automobile Engineering, Vol.213 (D), No.2, pp.145-159.

Heywood, J.B., 1988. Internal Combustion Engine Fundamental. McGraw-Hill Book Company.

History, BorgWarner Turbo & Emission Systems, URL: http://www.turbodriven.com/en/turbofacts/default.aspx [accessed 20th June 2007]

Kawaguchi J,. Adachi K., Kono S., Kawakami T., 1999, Development of the VFT (Variable Flow Turbocharger), Society of Automotive Engineers (SAE), paper No. 1999-01-1242

Moody, J.F., 1986. Variable Geometry Turbocharging with Electronic Control. Society of Automotive Engineers (SAE), paper No. 860107.

Nieuwstadt, V., Kolmanovsky, M.I., and Moraal, P., 2000, Coordinated EGR-VGT Control for Diesel Engines: an Experimental Comparison, Society of Automotive Engineers (SAE), paper 2000-01-0266.

Pesiridis, A., And Martinez-Botas, R., 2007. Experimental Evaluation of Active Flow Control Mixed-Flow Turbine for Automotive Turbocharger Application. Journal of Turbomachinery, Vol. 129, Issue 1, pp. 44-52.

Rajoo, S. & Martinez-Botas, R. 2007, Improving energy extraction from pulsating flow by active control of a turbocharger turbine. Society of Automotive Engineers (SAE), paper 2007-01-1557.

Shahed, S.M., 2005. The power of turbocharging. SAE online


magazine article, Edition September 2005, pp. 86-87, URL: www.aeu-online.org/automag [accessed 20th June 2007].

Stefanopoulou, A., Kolmanovsky, I., and Freudenberg, J.S., 2000 Control of Variable Geometry Turbocharged Engines for Reduced Emissions, IEEE Transactions on Control Systems Technology, Vol. 8(4).

Uchida, H., 2006. Trend of turbocharging technologies. R&D review of Toyota CRDL, Vol. 41, No. 3, Special Issue: Turbocharging Technologies, URL: http://www.tytlabs.co.jp/english/review/rev413e.html [accessed 20th June 2007]

Wallace, F., J., Howard, D., And Anderson, U., 1986, Variable Geometry Turbocharging–Optimization and Control Under Transient Conditions, In: 3rd International Conference on Turbocharging and Turbochargers, Institution of Mechanical Engineers, paper C98/86, pp. 227-240.

Walsham, B. E., 1990. Alternative Turbocharger Systems for the Automotive Diesel Engine. In: 4th International Conference on Turbocharging and Turbochargers, Institution of Mechanical Engineers, London, paper C405/036, pp. 39-50.

Watson, N., And Janota, M.S., 1982. Turbocharging the Internal Combustion Engine. London: The Maxmillan Press Ltd.

2

PD CONTROL PERFORMANCE FOR RATIO CONTROL OF AN EMDAP- CVT

Bambang Supriyo Kamarul Baharin Tawi

Hishamuddin Jamaluddin Sugeng Ariyono

2.1 INTRODUCTION Continuously Variable Transmissions (CVTs) have become very popular in automotive applications. Compared with a manual transmission or a stepped automatic transmission, CVT has wider range of transmission ratio coverage. This unique characteristic of CVT makes it possible for engine operating conditions to be adjusted accordingly to achieve its minimum fuel consumption or maximum engine performance.

Most of current CVTs employed hydraulic actuator. This kind of CVT needs continuous power from the engine to drive the hydraulic pump to generate line pressure. The line pressure is required to supply clamping force between the belt and pulley sheaves in order to maintain the desired transmission ratio and to prevent belt slip. In many operating situations, this continuous power consumption causes the major loss in the hydraulic CVT system, hence lowering CVT efficiency (Akehurst et al., 1999).

In addition, these CVTs are designed with single acting pulley mechanism. It means, that only one pulley sheave in each pair is movable, hence introducing belt misalignment. Long term application of this misalignment may damage the belt. This belt misalignment has been studied intensively in (Tawi, 1997).

24 Research on Vehicle Technologies (2008))

This chapter introduces an electro-mechanical dual acting pulley CVT (EMDAP CVT) system with two DC motors as its actuators. This actuator works only during transmission ratio changes, hence shortening actuator’s operation time and reducing energy loss. As compared to the Electro-mechanical CVT proposed in (Van de Meerakker et al., 2004) that employs one movable pulley sheave, the EMDAP CVT adopts two movable pulley sheaves in each shaft to eliminate belt misalignment. Each pair of movable sheaves in each shaft is driven by DC motor system. The primary motor is used for changing the CVT ratio, while the secondary one is used for preventing the belt from slipping.

PID (Proportional, Integral and Derivative) controller has been the basis in simple linear control systems for many years. The PID controller is a well-known technique for various industrial control applications, mainly due to its simple design, straightforward parameters’ tuning and robust performance. PID controllers are very common for motor control applications (Lin et al., 1994; Tang, 2001; Huang et al., 2008; Meenakshi, 2008). For motor position control applications, PD controller is the most popular.

In this chapter, a PD controller has been proposed for controlling the EMDAP CVT ratio. Section 2.2 introduces the background of CVT. Meanwhile, Section 2.3 provides a brief explanation on EMDAP CVT and the proposed controller. The results and discussions are given in Section 2.4 before the chapter is finally concluded in Section 2.5. 2.2 BACKGROUND OF CVT The basic of the CVT is actually similar to a variator which consists of a primary pulley, a secondary pulley and a metal belt connecting these two pulleys. The variator geometry schematic diagram is given in Figure 2.1. By assuming that the belt has a

PD Control Performance for Ratio Control of an EMDAP- CVT 25

fixed length, does not slip and moves at perfect circles with primary and secondary running radii Rp and Rs, respectively, the tangential velocities of both pulleys and belt will be the same.

Figure 2.1 Variator Geometry

The relationship between speed and running radii of the

variator can be given as follows:

ppss RR ωω = (2.1)

p

ssr ω

ω= (2.2)

p

sCVT R

Rr = (2.3)

Where, Rp - primary running radius Rs - secondary running radius ωp - angular speed of the primary shaft (input) ωs - angular speed of the secondary shaft (output) rs - speed ratio rCVT - CVT ratio


The implicit relationship between belt length L and running radii can be defined as:

L= (π+2θ) RP + (π-2θ) RS+2c Cos (θ) (2.4) RP = RS +c sin (θ ) (2.5)

Where,

L - belt length c - pulley center distance θ - half the increase in the wrapped angle on the

primary pulley. The relationship between running radii and pulley position

can be given as:

RP = Rp0 + (xP /tan (α)) (2.6) RS = Rs0 + (xS /tan (α)) (2.7)

Where

Rp0 - minimum primary running radius Rs0 - minimum secondary running radius α - pulley wedge angle (11º) xP - primary pulley position xS - secondary pulley position

By solving Equations (2.4) and (2.5), for L = 165 mm, and c= 645.68 mm, the relationship between running radii and belt wrapped angle can be plotted as is illustrated in Figure 2.2. By using Equation (2.3) and data from Figure 2.2, the relationship between running radii and CVT ratio can be plotted as shown in Figure 2.3.


Figure 2.2 Relationship between running radii and belt wrapped angle

Figure 2.3 Relationship between running radii and CVT ratio

2.3 EMDAP CVT The DC motor shaft of the EMDAP CVT is directly connected to a set of gear reducers and power screw mechanism to axially move the pulley sheaves. A Van Doorne’s metal pushing V-belt is placed between pulley sheaves, and runs on the surfaces of the sheaves.


Figure 2.4 Block diagram of EMDAP-CVT

This metal belt connects the primary and secondary pulleys to transmit the power and torque from the primary (input) to the secondary (output) shaft by means of friction between belt and pulley sheaves’ contacts (Kanehara et al., 1996; Fushimi et al., 1996). Both primary and secondary running radii determine the CVT ratio, which is indirectly measured via the pulley position sensors.

Both of the primary and secondary actuation system of the EMDAP CVT consists of a dc motor, two gear reducers, power screw mechanism and two movable metal pulley sheaves for clamping the metal belt. A spring disc is placed at the back of each secondary pulley sheave to keep the belt taut and to reduce excessive slip during ratio change.

The two gear reducers employed in the EMDAP-CVT system have an overall ratio of about 128:1. The gear reducer input is coupled to the DC motor shaft, whereas its output is connected to the power screw mechanism to move the pulley sheaves. The power screw mechanism converts every one rotational screw movement to about 2-millimeter axial movement. The block diagram of the EMDAP-CVT system can be seen in Figure 2.4.


2.3.1 Proposed Controller

This paper proposes PD controllers for both of the DC motor systems to track the reference CVT ratio. Based on Figure 2.3, the reference CVT ratio indirectly can be represented as a combination of its respective reference primary running radius and reference secondary running radius. To satisfy the control objective, the primary DC motor system tracks the reference primary running radius, whereas the secondary DC motor tracks the reference secondary running radius. The actual CVT ratio is calculated using actual values of primary and secondary running radii. The block diagram of the proposed controller is given in Figure 2.5.

Figure 2.5 Block diagram of the proposed controller

The transfer function of a PD controller has the following form:

GPD = Kp+ Kd ·s (2.8) Another form of PD transfer function can be represented as:

GPD = Kp(1+ Td ·s) (2.9) The value of Kd is described as :


Kd = Kp ·Td (2.10) Where

Kp - proportional gain Kd - derivative gain Td - derivative time constant

The Astrom-Hagglund method (Astrom and Hagglund,

1994) is used to determine initial gains of PD controller, namely, critical waveform oscillation period Tc and critical gain Kc. These two values can be obtained from a relay feedback experiment, which is a closed loop relay experiment involving a DC motor system as its plant, and a relay as its feedback controller. The block diagram for EMDAP CVT relay feedback experiment is shown in Figure 2.6. The critical gain resulted from relay feedback experiment can be defined as:

a

dK c π

4= (2.11)

Where

d - amplitude of the relay output a - amplitude of the waveform oscillation By using both values of Tc and Kc, the PD parameters can

now be calculated using Ziegler-Nichols formula (Hwang et al., 1999):

Kp = 0.6 Kc (2.12)

Td = 0.125 Tc (2.13)


(a)

(b)

Figure 2.6 Relay feedback controller for (a) primary and (b) secondary

DC motor systems

2.4. RESULTS AND DISCUSSIONS 2.4.1 Experimental Setup The EMDAP CVT experimental test rig is shown in Figure 2.7. The PD controller is implemented on this EMDAP CVT system. The test rig consists of EMDAP-CVT gear box, AC Motor unit, interfacing unit, Data Acquisition System (DAS) Card, Personal Computer (PC), and Power Supply unit. The schematic representation of the test rig is given in Figure 2.8.

EMDAP CVT gearbox has two DC motor systems for adjusting the pulley sheaves’ running radii, two position sensors for measuring pulley positions and two encoders for measuring shaft angular speeds. The AC motor unit acts as an engine, which drives the input shaft of the EMDAP CVT with a maximum speed of 80 rpm. The interfacing unit allows the DAS card to read sensors and to control the DC motors. The DAS card is fully controlled by computer via Matlab/Simulink Real Time Workshop package, which also performs the controller task for the EMDAP-CVT. The sampling time for this controller is set to 0.1 seconds.


Control performance of the PD controller is determined based on the overshoot, steady state error and trajectory errors.

Figure 2.7 Experimental Test rig

Figure 2.8 Schematic representation of test rig


2.4.2 Initial Parameters of PD Controllers

The EMDAP CVT ratio range is from 2.0 (underdrive) to 0.7 (overdrive). A step input corresponding to CVT ratio of 1.35 is used as input for relay feedback experiment. The amplitude of the relay controller is set to 10, since the input voltage in the range of -5 to +5 volts is needed to drive the DC motor system. Figure 2.9 shows the result of relay feedback experiment for primary DC motor system. The Vpd acts as reference input for primary pulley position, the Vpa represents the actual primary pulley position and Vrel is the relay controller output.

Figure 2.9 Result of relay feedback controller for primary DC motor system

Figure 2.10 Result of relay feedback controller for secondary DC motor system


Figure 2.10 shows the result of relay feedback experiment for secondary DC motor system. The Vsd acts as reference input for secondary pulley position, the Vsa represents the actual secondary pulley position and Vrel is the relay controller output.

From Figure 2.9, the values of Tc is 8.167 s, a and d are 0.423V and 10V, respectively. By using Equations (2.11), (2.12), (2.13), and (2.10), the Kp and Kd of the primary PD controller are 18.07 and 18.45, respectively. Similarly, for Figure 2.10, the values of Tc is 5.66s, a is 0.615V, and d is 10V. The Kp and Kd of the secondary PD controller are 12.42 and 8.79 respectively. 2.4.3 Fine Tuning of PD Controller The initial parameter values obtained from relay feedback experiment needs to be fine tuned. The tuning process is conducted by examining the output responses of primary and secondary pulley position sensor, VPPS and VSPS, respectively, when the ratio reference is up-shifted from 2 (underdrive) to 0.7 (overdrive) and down-shifted from 0.7 (overdrive) to 2 (underdrive). Pulley position sensors detect the current pulley positions of xP and xS. By applying equation (2.6) and (2.7), the current running radii of Rp, and Rs can be calculated. Finally, by applying equation (2.3) the current CVT ratio can be determined.

(a)


(b)

Figure 2.11 Response of the primary DC motor system: (a) overdrive (b) underdrive

(a)

(b)

Figure 2.12 Response of the primary DC motor system: (a) overdrive

(b) underdrive


The tuning process will only fine tune the proportional part of PD controller manually and leave the differential part unchanged. The output responses of the CVT ratio during fine tuning process for both primary and secondary PD controller are shown in Figures 2.11 and 2.12, respectively. The experimentally fine tuned values of proportional parts of PD controllers derived from Figures 2.11 and 2.12 are given in Table 2.1.

Table 2.1 Fine tuned proportional gains

CVT Ratio Primary DC Motor

System Secondary DC Motor

System Decreasing Kp=8 Kp=4 Increasing Kp=12 Kp=5

2.4.4 PD Controller Performance The fine tuned proportional gains and unchanged differential gains are selected as final gains for PD controllers of EMDAP CVT. The performances of these PD controllers are tested using square wave and sinusoidal input excitations. The results are given in Figures 2.13 and 2.14.

Figure 2.13 shows the responses of CVT ratio output during application of a square wave excitation input. It can be seen that time taken by the PD controllers to go from CVT ratio of 0.7 to 2 is about 13 seconds, whereas, the time taken to go from CVT ratio of 2 to 0.7 is about 15 seconds. In addition, PD controller has successfully eliminated the overshoot and minimized the steady state error in the ranges of – 0.0045 to 0.003 (underdrive), and – 0.0013 to 0.0034 (overdrive).

Figure 2.14 shows the responses of CVT ratio output during application of a sinusoidal wave excitation input. It can be seen that the PD controller has successfully tracked the input trajectory with an error in the range of -0.07 to 0.05.


(a)

(b)

(c)

(d)


(e)

Figure 2.13 (a) CVT ratio response (b) Underdrive settling time (c)

Overdrive settling time (d) Underdrive steady state error (e) Overdrive steady state error

(a)

(b)

Figure 2.14 (a) CVT ratio response (b) CVT ratio tracking error


2.5 CONCLUSION

The proposed control scheme of PD controller was examined to reduce the overshoot, steady state error and trajectory error for ratio control of EMDAP CVT system. The results of this work show that the application of Astrom-Hagglund method and Ziegler-Nichols formula is capable of providing a practical solution for obtaining initial parameters of the PD controllers. The application of experimentally fine tuning has shown a good performance for the PD controller. However, the authors believe that this is just a first start, with further works on PD auto tuning controller, better results can be obtained. In future it is suggested that an adaptive controller scheme should be considered in order to dynamically fine tune the controller parameters as a respond to the changes of plant’s parameters and loadings. REFERENCES Akerhurst, S., Vaughan, N.D. and Simner, D. 1999. An

Investigation into the Loss Mechanisms Associated with an Automotive Metal V-belt CVT. European Automotive Congress Vehicle Systems Technology for The Next Century, STA991407, Barcelona, pp. 342-350.

Astrom K.J. and Hagglund, T. 1984. Automatic Tuning of Simple regulators with Specifications on phase and amplitude margins, Automatica, 20, pp.645-651.

Fushimi, Y., Fujii, T. and Kanehara, S. 1996. A Numerical Approach to Analyse the Power Transmitting Mechanisms of a Metal Pushing V-Belt Type CVT. SAE Technical Paper Series 960720.

Huang, G. and Lee, S. 2008. PC-based PID speed control in DC motor. International Conference on Audio, Language and Image Processing, ICALIP 2008, pp.400 – 407.

Hwang, H.S. , Choi, J. N., Lee,W.H. and Kim, J.K. 1999. In A


Tuning for The PID Controller Utilizing Fuzzy Theory. ”IJCNN’99 International Joint Conference on Neural Network. Vol.4, pp. 2210-2215.

Kanehara, S., Fujii, T. and Oono, S. 1996. A Study of a Metal Pushing V-Belt Type CVT : Macroscopic Consideration for Coefficien of Friction between Belt and Pulley. SAE Paper No. 9636277.

Meenakshi, M. 2008. Microprocessor Based Digital PID Controller for Speed Control of D.C. Motor. First International Conference on Emerging Trends in Engineering and Technology, ICETET '08, pp.960 – 965.

Lin, P.H., Hwang, S. and Chou, J.1994. Comparison on fuzzy logic and PID controls for a DC motor position controller. Industry Applications Society Annual Meeting, 1994. Conference Record of the 1994 IEEE, Vol.3, pp.1930 - 1935 .

Tang, J. 2001. PID controller using the TMS320C31 DSK with online parameter adjustment for real-time DC motor speed and position control. IEEE International Symposium on Industrial Electronics, 2001, Proceedings ISIE 2001, Vol. 2, pp.786 – 791.

Tawi, K.B. 1997. Investigation of Belt Misalignment Effects on Metal Pushing V-Belt Continuously Variable Transmission. Ph.D. Thesis, Cranfield University.

Van de Meerakker, K., Rosielle, P., Bonsen, B. and Klaassen, T. 2004. Design of An Electromechanical Ratio and Clamping Force Actuator for a Metal V-Belt Type CVT. The 7th International Symposium on Advanced Vehicle Control (AVEC’04), Arhem, Netherland.

3

BRAKE SQUEAL CONTROL USING AN ACTIVE TECHNIQUE

Musa Mailah S.M Hashemi-Dehkordi

A.R. Abu Bakar

3.1 INTRODUCTION This chapter proposes a novel approach to suppress vibration that causes brake noise is proposed employing a closed-loop feedback control method using an Active Force Control (AFC) based strategy. It is used in conjunction with the classic proportional-integral-derivative (PID) scheme that is typically incorporated in the outermost positional control loop. The idea is to introduce an active element that dynamically compensates the disturbances through a control mechanism that takes into account the direct measurements and estimation of parameters in the AFC section. A disc brake model is considered and simulated taking into account a number of operating and loading conditions. Results clearly show the superiority of the proposed AFC-based scheme compared to the pure PID counterpart in suppressing the vibration and hence the brake noise.

Brake is one of the most important safety components in any automotive vehicle and is almost indispensable. The brake system of an automobile typically consists of the contact metallic solids rubbing against each other, which frequently generates undesirable noise and vibrations. This in turn causes discomfort to the passengers and adversely affects their perceptions of the quality and reliability of the vehicle. Thus, noise generation and


suppression have become an important factor to be considered in the design and manufacture of brake components. Indeed, as noted by Abendroth and Wernitz (2000), a large number of manufacturers of brake pad materials spend up to 50% of their engineering costs on issues related to noise, vibration and harshness. Brake noise vibration phenomena are described by a number of terminologies that are sometimes interchangeably used such as squeal, groom, chatter, judder, moan and hum (Kinkaid et al., 2003). Even to this day, there is no precise or conclusive definition of brake squeal that has gained complete acceptance. It is also worth mentioning that since in a vehicle with disc brakes installed at the front wheels while drum brakes at the back wheels, around 70% of the braking action occurs at the front wheels. Thus, it is expected that most of the noise and squeal is coming from the front disc brake system. The main cause of brake noise is due to the effect of negative damping (Shin et al., 2002), in which the friction coefficient reduces with the increase of relative velocity of the pad and the disk. The other factor maybe attributed to mode coupling (Chen et al., 2006).

In literature, there are three major methods to study and reduce brake squeal, namely through mathematical modeling, experimental and finite element methods. A recent research study has been carried out for reducing brake noise using finite element (FE) can be found in (Dai and Lim, 2007). They developed a dynamic FE model of the brake system, and based on their analysis, the pad design changes can be made in the FE model to determine the potential improvements in the dynamic stability of the system and also in noise reduction. Wagner et al. proposed a new mathematical rotor based model of a brake system that is suitable for noise analysis (Wagner et al., 2007). A brief description of the previous mathematical models that have been developed by other researchers were explicitly outlined in their study. Besides, there is also an active control method known as dither control which makes use of high frequency disturbance signal for the suppression of the automotive disc brake squeal.

Brake squeal control using an active technique 43

Through this scheme, the dither signal stabilizes friction induced self-oscillations in the disc brake using a harmonic vibration, with a frequency higher than the squeal frequency generated from a stack of piezoelectric elements placed in the caliper piston of the brake system. The resulting control vibration was not heard from the brake system if an ultrasonic control signal was activated. This system assumes an open loop control mode in which there is no requirement to detect the presence of squeal and is much simpler in design than the feedback control (Grag, 2000). This paper presents a closed loop control employing system Active Force Control (AFC) with PID element applied to a brake model described in (Hoffmann et al., 2005) in order to suppress the brake noise and squeal. The main advantage of the AFC technique is its ability to reject disturbances that are applied to the system through appropriate manipulation of the selected parameters. In addition, the technique requires much less computational burden and has been successfully demonstrated to be readily implemented in real-time. AFC as first proposed by Hewitt and Burdess (1981) is very robust and effective in controlling a robot arm. Mailah has successfully demonstrated the application of the technique to include many other dynamical systems with the incorporation of artificial intelligence (AI) methods (Mailah, 1998; Mailah and Rahim, 2000). 3.2 The Disc Brake Model A disc brake system assuming a two degree of freedom model based on the one described in (Hoffmann et al., 2005) is considered in the study. The model consists of a conveyor belt with constant velocity νB that is pushed with a constant normal force FN against a block modelled as a block of mass m. Figure 3.1 shows that the model is just a single-point mass sliding over a conveyor belt and there are two linear springs k1 and k2 parallel and normal to the belt surface with the latter regarded as the physical contact stiffness between the objects in relative sliding motion. In addition,


there is another linear spring k mounted at oblique angle of 45° constituting the off-diagonal terms in the model’s stiffness matrix. For the friction component, a coulomb model is assumed such that FT = µ FN, where µ is the coefficient of kinetic friction usually taken to be constant. FN is a normal force and since the normal force at the friction interface is linearly related to the vertical displacement x2 of the mass then the resulting friction will become FF = µ k2 x2. Assuming that the mass of the conveyor belt system is larger than the mass block, it implies that the vibration of the belt does not show any changes due to its inertia.

Figure 3.1 Two degree of freedom disc brake model The matrix form of the equation of motion can be expressed as (Hoffmann et al., 2005):

−=

+−

−++

+

N

xk

x

xk

kk

kkk

x

xc

c

x

xm

m 22

2

1

2

1

.

2

.

1

2

1

2..

..

1

22

220

00

0

µ (3.1)


3.3 CONTROL STRATEGY Upon acquiring the model of the disc brake and its related equation of motion, it is required to control the vibration of the mass with respect to the vertical direction, x2 considering an actuator that produces a force F in parallel with the preload force N. Thus, Equation (3.1) will be written as:

+−

=

+−

−++

+

FN

xk

x

xk

kk

kkk

x

xc

c

x

x

m

m 22

2

1

2

1

.

2

.

1

2

1

..

..

2

1

22

220

0

0

0 µ (3.2)

A robust control strategy is proposed here employing an Active Force Control (AFC) based scheme that is used in conjunction with the conventional PID controller. The PID controller was first tuned with Ziegler-Nichol’s method for good performance and later the AFC part was incorporated into the system to provide the compensation of the disturbances. Figure 3.2 shows the AFC scheme applied to a dynamic translation system (disc brake). AFC scheme is shown to be very effective provided the actuated force and body acceleration are accurately measured and at the same time the estimated mass property approximated (Hewit and Burdess, 1981; Mailah, 1998). The essential AFC equation can be related to the computation of the estimated disturbance Fd as follows:

aMFFd '.−= (3.3)

where F is the measured actuating force, M’ is the estimated mass and a is the measured linear acceleration. This parameter is then fed back through a suitable inverse transfer function of the actuator to be summed up with the PID control signal. The theoretical analysis including stability of the proposed AFC method has been sufficiently described in (Burdess and Hewit, 1986). A number of


methods to estimate the mass has been proposed in previous studies such as through the use of artificial intelligence (AI) and crude approximation techniques (Hewit and Burdess, 1981; Mailah, 1998; Mailah and Rahim, 2000). In this study, the use of crude approximation method to approximate the estimated mass is deemed sufficient. The main challenge of the AFC method is to acquire appropriate estimation of the mass needed to compute the disturbance Fd in the feedback loop. A conventional PID that is used with the AFC scheme can be typically represented by the following equation:

Gc(s) = Kp + Ki/s + Kd s (3.4) where Kp, Ki and Kd are the proportional, integral and derivative gains respectively.

Figure 3.2 Schematic diagram of AFC strategy


3.4 SIMULATION MATLAB, Simulink and Control System Toolbox (CST) software were used to simulate the brake model with the controllers. The actuator is assumed to be a linear type with a suitable constant gain. It provides the necessary external energy to suppress vibration in the model. The parameters used in this study were taken from the previous research (Hoffmann et al., 2002). However, some of them need to be modified to suit the application in the simulation. The detailed parameters are as follows: Minimal disc brake model parameters:

• Body mass, m = 0.7 kg, • Spring stiffness, k = 10 N/m, k1 = 11 N/m, k2 = 20 N/m • Damping coefficient, c1 = 0.4 Ns/m, c2 = 0.4 Ns/m • Friction coefficient, µ = 0.3 • Normal preload, N = 5 N

Actuator: • Actuator gain, Q = 0.5

Reference value: • Reference input = 0.00 m (i.e. no vibration)

Disturbances: • Magnitude of step function = 3 N • Amplitude of sinusoidal function = 6 N (12 N), Frequency =

1.5 Hz (2.5 kHz) In this work, two types of disturbances, namely the step and sinusoidal disturbances are deliberately introduced to the disc brake system to evaluate the robustness of the system. The Simulink diagram of the passive disc brake system model is shown in Figure 3.3. The schematic block diagram was constructed from Equation (3.1).


Figure 3.3 A passive brake system model

In order to have an active disc brake system, an actuator force for compensating the disturbance force is required, and the actuator force is controlled by a PID controller which typically involves a negative feedback loop. Hence, there are two inputs to the dynamic disc brake system which is the sinusoidal disturbance and the actuator force input. Figure 3.4 shows a configuration of the active disc brake system.

Figure 3.4 A PID active brake control scheme


To achieve better overall performance of the brake system, AFC is ‘added’ to the PID controller, The AFC Simulink blocks includes the estimated mass, parameter 1/Q and the percentage of AFC gain. The input to the AFC control is the mass acceleration and the output is summed with the PID controller output and then multiplied with the actuator gain which finally generated the actuator force. In order to get the effective results using this method, it is required to have a suitable mass estimation combined with the best tuning of the PID controller gains. Also a memory block is used to eliminate the algebraic loops and algebraic variable problems. Figure 3.5 shows an AFC scheme that assumes a step input as its disturbance.

Figure 3.5 Proposed AFC model

Note that the AFC Simulink model with sinusoidal disturbance is the same as those shown in Figure 3.5 but with different disturbance input. To tune the PID controller, we use the Ziegler-Nichol’s method and the results are tabulated as shown in Table 3.1.


Table 3.1 PID parameters tuned using Ziegler-Nichol’s method

PID Kp Ki Kd Ti Td

Gain 0.6 0.857 0.105 0.7 0.17

The estimated mass for the AFC loop was obtained by trial and error (crude approximation method) in which a suitable value was easily found to be 1.4 kg, and the percentage of AFC used is 100% implying that the AFC loop employs full AFC implementation. 5.1 RESULTS AND DISCUSSION The simulation was executed for a period of 5 s after tuning the PID control system and obtaining suitable values for other relevant parameters. It is usual that the brake process does not take more than that duration for the purpose of observing and analysing the response. At first, the step disturbance is applied and then the simulation was performed without using AFC and only the pure PID controller was applied. The result of this process can be seen in Figure 3.6.

It can be observed that the vibration that may result in producing noise or squeal is relatively high (a peak of more then 0.75 m in amplitude in some regions). Later, the simulation was repeated but this time around, the AFC mechanism (plus PID) was activated. A switch can be employed in the overall scheme to switch from the PID only mode to the AFC plus PID scheme. The result of the scheme is superimposed onto previous graph obtained as shown in Figure 3.6. It is evident that the vibration is virtually reduced to a very low level, implying that squealing can be almost if not totally avoided through the AFC-based scheme.


Figure 3.6 Response of the brake model with a step input disturbance for systems with PID only and PID+ AFC

A closed-up view of the graphical result of the PID+AFC scheme is depicted in Figure 3.7 for further analysis.

Figure 3.7 A close-up view of the performance of the PID+AFC controller


It clearly shows the superiority of the scheme in rejecting the vibration of the system. The maximum amplitude of the vibration was less then 0.06 m, which is very much lower compared to the brake system operated with a PID only controller. Responses in frequency domain were also carried out and the results are shown in Figure 3.8. The PID controller obviously produces amplitudes with much higher peaks compared to the AFC-based scheme, implying that the vibration is indeed reduced significantly to about 1 unit in magnitude which is in stark contrast to the PID only scheme (almost 45 units). Figures 3.9 to 3.12 show results obtained through the application of harmonic or sinusoidal disturbances with two different sets of amplitudes and frequencies. Again, the results fully complement the findings from the previous disturbance condition and thereby verifying the superiority of the AFC-based scheme compared to the PID system in suppressing the vibration effect. It should be noted that when the high frequency disturbance was applied, the estimated mass in the AFC loop was set to 0.5 kg.

(a) (b)

Figure 3.8 Responses in frequency domain for

(a) PID+AFC (b) PID schemes


Figure 3.9 Response of the brake model with sinusoidal disturbance for systems with PID only and PID+AFC (amplitude = 6 N,

frequency = 1.5 Hz)

(a) (b)

Figure 3.10 Responses in frequency domain for (a) PID+AFC (b) PID

schemes (amplitude = 6 N, frequency = 1.5 Hz)


Figure 3.11 Response of the brake model with sinusoidal disturbance for systems with PID only and PID+AFC amplitude = 12 N, frequency = 2.5 kHz)

(a) (b)

Figure 3.12 Responses in frequency domain for (a) PID+AFC

(b) PID schemes (amplitude = 12 N, frequency = 2.5 kHz)

Since brake noise is typically inherited in a brake system, it


will always exist even in the absence of any external disturbances due to presence of the constant preload force, N as described by Equation (3.1). To verify this fact, the simulation was executed without any disturbances and the result can be seen in Figure 3.13. The result is almost similar to that obtained when a high frequency sinusoidal disturbance was applied to the system.

Again, the AFC-based scheme is the more effective controller then the PID only counterpart in reducing the undesirable vibration and hence the brake noise or squeal. It is useful to note that the ‘robust’ term used throughout the undertaken research study is exclusively meant to describe the conditions specifically designed for the proposed works. Further investigation is needed to ascertain its robust performance when other more critical operating and loading conditions were considered.

Figure 3.13 Response of the brake model without any external disturbance for systems with PID only and PID+ AFC

Figures 3.14 and 3.15 respectively show the results of the

system without the application of any external forces and the other, subject to an increase in the friction coefficient and the preload force N.


Figure 3.14 Response of the brake model without any external disturbance and a friction coefficient of 0.6 for systems with PID only and PID+ AFC

Figure 3.15 Response of the brake model without any external disturbance and a preload force of 10N for systems with PID only and PID+ AFC


3.6 CONCLUSION From the undertaken research, a novel AFC-based scheme has been developed to effectively suppress the vibration and noise (squeal) emanating from a disc brake system. It is obvious that when a pure PID controller is applied to the brake system, vibration and noise are in fact reduced but still a noticeable amount of them remain. However, upon applying the AFC-based technique, the vibration and noise (squeal) are significantly reduced and approaching zero datum. This implies that the proposed strategy is robust and effective in countering the undesirable effects. Future works may include the study of the system behaviours considering other different operating and loading settings. An experimental rig to verify the theoretical concept is also worthwhile exploring and maybe a viable option. REFERENCES Abendroth, H. and Wernitz, B. 2000. The Integrated Test Concept:

Dyno-vehicle, Performance-noise”, Technical Report 2000-01-2774, SAE, Warrndale, PA.

Burdess, J.S. and Hewit, J.R. 1986. An Active Method for the Control of Mechanical Systems in the Presence of Unmeasurable Forcing. Mechanism and Machine Theory, 21(5): 393-400.

Chen, F., Tan, C.A. and Quaglia, R.L. 2006. Disk Brake Squeal: Mechanism, Analysis, Evaluation and Reduction/Prevention. SAE Paper.

Dai, Y. and Lim, T.C. 2007. Suppression of Brake Squeal Noise Applying Finite Element Brake and Pad Model Enhanced by Spectral-based Assurance Criteria. Applied Acoust. doi: 10.1016 / j.apacoust. 2006.09.010

Grag, A. 2000. Active Control of Automotive Disc Brake Rotor Squeal Using Dither. Master Thesis, Georgia Institute of


Technology, 2000. Hewit, J.R. and Burdess, J.S. 1981. Fast Dynamic Decoupled

Control for Robotics Using Active Force Control. Trans. on Mechanism and Machine Theory, 16(5): 535-542.

Hoffmann, N., Fischer, M., Allgaier, R. and Gaul, L. 2002. A Minimal Model for Studying Properties of the Mode-coupling Type Instability in Friction Induced Oscillations”, Mechanics Research Communications, 29: 197-205.

Hoffmann, N., Wagner, N. and Gaul, L. 2005. Quenching Mode-coupling Friction-induced Instability Using High-frequency Dither. Journal of Sound and Vibration, 279: 471–480.

Kinkaid, N.M., O’Reilly, O.M. and Papadopoulos, P. 2003. Review: Automotive Disc Brake Squeal. Journal of Sound and Vibration, 267: 105-166.

Mailah, M. 1998. Intelligent Active Force Control of A Rigid Robot Arm Using Neural Network and Iterative Learning Algorithms. Ph.D Thesis, University of Dundee, UK.

Mailah, M. and Rahim, N.I.A. 2000. Intelligent Active Force Control of A Robot Arm Using Fuzzy Logic. Proceedings of IEEE International Conference on Intelligent Systems and Technologies, Kuala Lumpur, Malaysia, Vol.2, pp. 291-296.

Shin, K., Brennan, M.J. and Oh, J.-E. and Harris, C.J. 2002. Analysis of Disk Brake Noise Using A Two-degree-of-freedom Model. Journal of Sound and Vibration, 254(5): 837–848.

Wagner, U.V., Hochlenert, D. and Hagedorn, P. 2007. Minimal Models for Disk Brake Squeal. Journal of Sound and Vibration, 302: 527-539.

4

REDUCTION OF EMISSIONS IN DIRECT INJECTION DIESEL ENGINE USING HIGH TURBULENCE COMBUSTION

WITH EXHAUST GAS RECIRCULATION AND PILOT

INJECTION Wira Jazair

Yoshiyuki Kidoguchi Kei Miwa

4.1 INTRODUCTION Combustion in a compression ignition (CI) engine or diesel engine is quite different from that in a spark ignition (SI) engine. Whereas combustion in an SI engine is essentially a flame front moving through a homogeneous mixture, combustion in CI engine is an unsteady process occurring simultaneously at many spots in a very non-homogenous mixture at a rate controlled by fuel injection. Engine torque and power output controlled by the amount of fuel injected per cycle. In diesel engine, fuel is injected into the cylinders late in the compression stroke by a fuel injector. Injection time is usually about 20o of crankshaft rotation, starting at about 15o before top dead center (TDC) and ending about 5o after TDC. After injection the fuel must go through a series of events such as atomization, vaporization, mixing, self-ignition and combustion. These events or the combination of these events need to be controlled to assure the proper combustion process and emit lower exhaust emissions.


There are many stategies to control the diesel combustion.

The research development on high-pressure injection, multiple injection, EGR (Exhaust Gas Recirculation) and optimization of combustion chamber geometry are some of mechanisms that can improve combustion process. Furthermore, aftertreatment system such as DPF and de-NOx catalyst are also been developed. Some petroleum companies are also struggling to produce reformulated fuels for improving the exhaust emissions of vehicles.

Among emission control methods stated above, the fuel-air mixing control technique by optimization of combustion chamber geometry has gain special attension for years (Kidoguchi et al., 1999, 2003, 2005). Piston geometry for high turbulence combustion (named as R35S) compared with standard toroidal combustion chamber (STD) are shown in Figure 4.1. These studies suggested that strong swirl and squish obtained by combustion chamber geometry produce fuel-rich and high turbulence combustion, which results in the reduction of NOx and particulate emissions. In addition, through computational calculation, it can be said that the high squish combustion chamber with a central pip is effective to keep the combustion mixture under the squish lip until the end of combustion and the combustion region forms rich and highly turbulent atmosphere. This kind of mixture distribution

20.7

R35S

STD

Spray

φ102

φ56.7

φ58.8

φ36.0

19.7

20.7

R35S

STD

Spray

φ102

φ56.7

φ58.8

φ36.0

19.7

Figure 4.1 Combustion chamber geometry

Reduction Of Emissions In Direct Injection Diesel Engine 61

tends to reduce initial burning, resulting in restraint of NOx emission while keeping low particulate emission.

Multiple injection is one of important technique to improve combustion as well as exhaust emission in diesel engines (Cheolwoong et al., 2004, Mallmo et al., 2002). These studies reported that pilot injection gives lower noise and NOx emission due to lower initial rate of heat release of main combustion. However, long diffusion combustion leads to higher particulate emission. This problem can be solve by selecting exact injection timing for both pilot and main injection to prevent particulate emision from becoming worse. Another way is by injecting post injection with proper quantity and timing. This post injection pulse can help burn off soot produced during main combustion. Applying EGR can also reduce overall combustion temperature and reduce NOx emission formation inside combustion cylinder. However, many of those related studies concluded that higher EGR rate can increase particulate emission particularly solid organic fraction (SOLID) at high load and soluble organic fraction (SOF) at low and medium engine load. It is also been reported that the engine thermal efficiency was reduced as EGR rate increased.

Those three important strategies of controlling combustion process and reducing certain exhaust emissions of diesel engines stated above (air-fuel mixing control, multiple fuel injection and EGR) have their own advantageous and disadvantageous. It would be interesting to investigate the combustion performance and exhaust emission of combinations those three strategies applied to a DI diesel engine. In this study common rail type fuel injection system has been used in order to produce high injection pressure. Both STD and R35S chambers used the same 5-hole injection nozzle with hole diameter of 0.21mm and spray angle of 145°. Injection pressure is 75MPa and swirl ratio is 3.6 in the case of R35S chamber. On the other hand, injection pressure is 100MPa and swirl ratio is 2.2 for the STD chamber (Goda et al., 2003). JIS#2 diesel fuel was used as tested fuel.


4.2 COMPARISON WITH THE STANDARD TOROIDAL COMBUSTION

Figure 4.2 shows cylinder pressure P, rate of heat release ⊿Q, and nozzle needle lift N.L. against crank angle θ under the condition of mean effective pressure of Pe=0.7MPa and Pe=0.2MPa with injection timing of θi = TDC for both R35S and STD combustion chambers. At Pe=0.7MPa, there is little change of ignition delay period due to combustion chamber geometry differences, however, the initial rate of heat release for STD chamber is higher. This is because higher injection pressure of STD chamber than R35S chamber can possibly increase the quantity of fuel in the air-fuel mixing during ignition delay. On the other hand, R35S chamber has higher rate of heat release at diffusion combustion stage. R35S chamber enhanced combustion at diffusion combustion stage due to intensified reverse squish flow and high turbulence caused by the squish lip that promote the mixing of unburned fuel with surplus air around squish area. At Pe = 0.2MPa, maximum rate of heat release for R35S chamber is higher than STD chamber, however, STD chamber has sharper slope of initial rate of heat release than R35S chamber.

Figure 4.2 Comparison of combustion process

0

2

4

6

8

Pe = 0.7MPa

⊿Q

J/

deg

θi = TDC

p

MP

a

⊿Q

p

-30 -15 TDC 15 30θ oATDC

N.L.

Pe = 0.2MPa

0

50

100

150

200

R35SSTDR35S

STD

-15 TDC 15 30 45


Figure 4.3 Comparison of gas velocity and fuel mass fraction between R35S and STD chambers

5degATDC

10degATDC

15degATDC

20degATDC

30degATDC

40degATDC

Fuel mass fraction

R35S

Gas velocity

0

0.02

0

6e+3 cm/s

0

6e+3 cm/s

Gas velocity

0

0.02

Fuel mass fraction

STD


Figure 4.3 compares gas velocity and fuel mass fraction at crank angle from 5oATDC to 40oATDC. These results are calculated from using CFD (Computational Fluid Dynamics) code KIVA-3VR2 program. As seen from the air motion in the STD chamber, burned gas flows from the outer edge of the cavity to the center along the bottom wall of the chamber. However, for the R35S chamber, burned gas flows from beneath the squish lip to the cavity bottom along the outer wall of the cavity. The air motion then turns upward along the central pip in the chamber. The R35S chamber tends to keep the rich mixture under the squish lip for a long time. This can control initial combustion and keep small region with high heat release.

Figure 4.4 shows combustion and exhaust emissions characteristics for R35S and STD chambers with injection timing θi =TDC against engine loads. Upper part of the figure is the combustion characteristics showing rate of maximum pressure increment ⊿pmax, maximum cylinder pressure pmax, combustion duration ⊿Qb, maximum rate of heat release ⊿Qmax, initial combustion ratio Qp/Qt and ignition delay period τ. Down part of the figure shows exhaust emission characteristics comprising NOx concentration, THC concentration, smoke concentration S, fuel consumption be, particulate emission PART, solid organic fraction SOLID, soluble organic fraction SOF, smoke concentration Opacity and CO concentration.

Regarding combustion characteristics, ⊿Qmax for mean effective pressure of Pe=0.7MPa to Pe=0.35MPa , STD chamber is higher than R35S chamber and ⊿pmax, pmax, Qp/Qt also give similar results. This is due to differences in injection pressure. At lower mean effective pressure Pe=0.2MPa and Pe=0MPa, there is no change on combustion performance due to differences of combustion chamber geometry. It is important to be noted that combustion duration for R35S chamber is shorter at high load.

As for exhaust emission characteristics, at Pe=0.7MPa, R35S chamber has same level of NOx concentration with STD chamber.


Strong airflow inside R35S chamber can caused a promoted

air-fuel mixing during ignition delay period and predicted to

Figure 4.4 Effect of high turbulence mixing on combustion and emission characteristics

0 0.2 0.4 0.6200

250

300

0

1

0

500

1000

400

800

1200

0

50

100

0

50

0

50

4

6

8

0.4

0.8

1.2

100

200

300

400

50

100

0 0.2 0.4 0.6 0.80

4

8

0 0.2 0.4 0.6 0.80

2000

0 0.2 0.4 0.60

10

20

30

0

10

be g

/kW

hS

Bos

chT

HC

ppm

C

STD

θi = TDC

Pe MPa P

e MPa

Pe MPaP

e MPa

NO

x ppm

PA

RT

mg/

m3

SO

LID

mg/

m3

SO

Fm

g/m

3

p max

MP

a

⊿p m

ax

MP

a/de

g

⊿Q

max

J/de

g

R35S

STD

R35S

τ d

egQ

p/Q

t %

θi = TDC

O

paci

ty%

CO

ppm

⊿θ b

deg


produce higher NOx concentration. However, lower injection pressure suppressed the formation of NOx emission. SOLID and particulate emissions for R35S chamber at Pe=0.7Mpa is lower than STD chamber. In STD chamber, combustion gases that include unburned air-fuel mixture intrude the squish area at early timing and freeze due to lower surrounding temperature. This might be the reason of high SOLID emission of using STD chamber. On the other hand, R35S chamber keep the combustion occurs under the squish lip with intensified reverse flow and high squish for a long period time (Yang et al., 1999). Intensified reverse squish and high turbulence promote the mixing of unburned fuel or combustion gas with surplus air around squish area and promoting soot to be oxidized. Therefore, R35S chamber has an effect of controlling smoke and particulate emissions by its high turbulence combustion and zero smoke combustion can be realize at high load engine condition. High THC emission of R35S chamber at Pe=0MPa is caused by misfire. This misfire occurs when low quantity of fuel that is injected at low load is further diluted by strong airflow to produced air-fuel mixture that has lower concentration than combustible limit.

Figure 4.5 Effect of injection timing on combustion process

0

2

4

6

8

R35S

⊿Q

J/

deg

Pe = 0.7MPa

p

MP

a

7.5o

2.5o

5o

TDCθ

i=-2.5oATDC

⊿Q

p

-30 -15 TDC 15 30θ oATDC

N.L.

2.5oTDC

STD

0

50

100

150

2005o

θi=-2.5oATDC

-15 TDC 15 30 45


Figure 4.5 shows histories of combustion process against injection timing θi at Pe=0.7MPa for both R35S and STD chambers. R35S chamber can produce high rate of heat release at diffusion combustion stage. Maximum rate of heat release increases as injection timing is retarded for STD combustion chambers. R35S combustion chamber give lower maximum rate of heat release at injection timing θi = 0 deg. In addition, slope of initial rate of heat release for R35S is rather lenient.

Figure 4.6 shows combustion performance and exhaust emission characteristics against injection timing θi at Pe=0.7MPa for both R35S and STD chambers.

Regarding combustion performance, as injection timing retarded, ⊿Qmax and ⊿pmax for STD chamber are increase. On the other hand, the increment of maximum rate of heat release for R35S chamber is suppressed. Maximum rate of pressure ⊿pmax for R35S chamber is reduce as injection timing retarded up to θi = 0.0oATDC and become unchanged from afterwards injection timing. ⊿Qb for R35S chamber shows shorter than STD chamber for any injection timing. Furthermore, Qp/Qt for R35S is lower and this mean that diffusion combustion for R35S chamber has been promoted to have short combustion period. Usually, initial rate of heat release will be promoted if there is higher injection pressure and produce shorter combustion period, however, R35S chamber can have high speed combustion without dependence on high pressure fuel injection.

As for exhaust emission characteristics, R35S chamber shows almost same level of NOx emission as STD chamber with injection timing earlier than θi = 5oATDC. This is because R35S chamber has lower injection pressure and moreover, stratified rich combustion may restrain NOx formation. However, there is increment of NOx concentration at injection timing θi = 7.5oATDC. This is because ignition delay period become longer as injection timing is retarded and produce higher amount of air-fuel mixture that will promote initial combustion. Smoke as well as particulate concentration for R35S chamber is suppressed at retarded injection


timing. Here, the using of R35S chamber with retarded injection timing can be utilized for simultaneous reduction of NOx and particulate emissions. However, retarded injection timing caused higher fuel consumption. The better solutions for NOx reduction without compromise with higher fuel consumption need to be found.

Figure 4.6 Effect of injection timing on combustion and emission

-4 -2 TDC 2 4 6200

250

300

0

1

0

500

1000

400

800

1200

0

50

100

0

50

0

50

4

6

8

0.4

0.8

1.2

100

200

300

400

50

100

-4 -2 TDC 2 4 6 80

4

8

-4 -2 TDC 2 4 6 80

2000

-4 -2 TDC 2 4 60

10

20

30

0

10

be g

/kW

hS

Bos

chT

HC

ppm

C

STD

Pe = 0.7MPa

θi oATDC θ

i oATDC

θi oATDCθ

i oATDC

NO

x ppm

PA

RT

mg/

m3

SO

LID

mg/

m3

SO

Fm

g/m

3

p max

MP

a

⊿p m

ax

MP

a/de

g

⊿Q

ma

x

J/de

g

R35S

STD

R35S

τ d

egQ

p/Q

t %

Pe = 0.7MPa

O

paci

ty%

CO

ppm

⊿θ b d

eg


4.3 EFFECT OF HIGH TURBULENCE COMBUSTION WITH PILOT INJECTION

The utilization of pilot injection together with high turbulence combustion chamber R35S chamber has been made. The effect of pilot injection on high turbulence combustion is investigated. 4.3.1 Effect At High Load Figure 4.7 shows combustion process against pilot injection timing θp at mean effective pressure Pe=0.7MPa for STD chamber with pilot injection. Here, main injection θi is at TDC. In this figure, qp/qt represent the ratio of pilot fuel delivery qp to total fuel delivery qt. Here qp/qt =12%.

As for rate of heat release ⊿Q, pilot injection near to TDC, θp= –15oATDC and θp =–30oATDC has sudden increase of pilot combustion. Around TDC, due to high in-cylinder temperature, air-fuel mixture produced by high turbulence of R35S chamber is ignited before it is being diluted and produces rapid pilot combustion. Therefore, as θp is further advanced, pilot combustion become lenient. Especially at θp=–60oATDC, in-cylinder temperature is lower during pilot injection. This low in-cylinder temperature cause a longer ignition delay period. Air-fuel mixture is diluted by turbulence during this ignition delay period and produces lenient pilot combustion.

Figure 4.8 shows combustion process against pilot injection timing θp at mean effective pressure Pe=0.7MPa for R35S chamber with pilot injection. Main injection θi is at TDC with injection pressure 75MPa. Similar to STD chamber shown in Figure 4.7, pilot injection near to TDC, θp=–15oATDC and θp = –30oATDC has sudden increase of pilot combustion while at θp= –60oATDC lenient pilot combustion is obtained.

Enhanced pilot combustion is obtained for both combustion chambers as pilot injection timing approached close to the main injection timing. This is significant for R35S chamber with higher


Figure 4.7 Effect of pilot injection timing on STD chamber combustion process

Figure 4.8 Effect of pilot injection timing on R35S chamber combustion process

0

2

4

6

8

0

50

100

150

200

-30 -15 TDC 15 30 45

-15

qp/q

t=12%

-30

θp=-60oATDC

w/o pilot

STD 1800rpmP

e=0.7MPa

Pinj

=100MPaθ

i=TDC

p M

Pa

-45⊿Q

⊿Q

J/

deg

p

θ oATDC

N.L.

0

2

4

6

8

0

50

100

150

200

-30 -15 TDC 15 30 45

-60

qp/q

t=12%

-30

θp=-15oATDC

w/o pilot

R35S 1800rpmP

e=0.7MPa

Pinj

=75MPaθ

i=TDC

p M

Pa

-45⊿Q

⊿Q

J/

deg

p

θ oATDC

N.L.


maximum pilot heat release compared with STD chamber. This might due to the effect of high turbulence air flow produced by R35S chamber. Figure 4.9 shows combustion performance and exhaust emission characteristics against θp for Figure 4.7. Here, θi=TDC and qp/qt=12%. ⊿Qm max in the figure is maximum rate of heat release for main injection, ⊿Qp max is maximum rate of heat release for pilot injection, ⊿pm max is maximum rate of pressure increment for main injection, ⊿pp max is maximum rate of pressure increment for pilot injection and Qpilot/Qt is rate of pilot combustion. Regarding combustion performance, rapid rise of pilot combustion can be suppressed by further advanced of pilot injection timing resulting reduction of ⊿Qp max and ⊿pp max. As θp is advanced, Qpilot/Qt is also reduced. As for pilot injection that advanced far from main injection, most of pilot fuel remain in the combustion chamber as air-fuel mixture and will ignite with main fuel injection. NOx concentration increased as pilot injection timing retarded. Smoke and PART emissions deteriorate at θp=–30oATDC. This is because some portion of fuel is injected outside bowl shape combustion chamber. Fuel fraction that burn inside the top clearance has a higher probability to quench due to lower surrounding temperature and produce more Smoke and PART emissions. Smoke and PART emissions reduce as pilot injection timing advanced surpasses θp=–30oATDC. As injection timing further advanced, injected fuel has longer mixing period and to be further diluted before main injection occurs. This can reduce over rich region inside combustion cylinder during main fuel injection. Figure 4.10 shows combustion performance and exhaust emission characteristics against θp for Figure 4.8. θi=TDC and qp/qt=12%. Regarding combustion performance, rapid rise of pilot combustion can be suppressed by further advanced of pilot injection timing resulting reduction of ⊿Qp max and ⊿pp max. As θp is advanced, Qpilot/Qt is also reduced. As for pilot injection that


advanced far from main injection, most of pilot fuel remain in the combustion chamber as air-fuel mixture and will ignite with main fuel injection. There is no change on NOx concentration as pilot injection timing changed. Pilot injection can reduce the initial heat release of main combustion and has positive effect of reducing NOx emission. However, higher temperature produced from pilot combustion itself contributes to the formation of NOx emission. Smoke and PART emissions deteriorate as pilot injection started near to TDC. Pilot injection near to TDC produces higher temperature at particular crank angle. Therefore ignition delay period for main combustion become shorter and produce main combustion that is mainly comprises of diffusion combustion. Furthermore, during main injection period, fuel is injected in the space where pilot combustion is already commenced and oxygen concentration is low at that particular area. Meanwhile, these Smoke and PART emissions can be reduced by advanced the pilot injection timing. At θp= –30oATDC or further advanced timing, it is likely that mixing of air and pilot injected fuel is promoted by intensified airflow and turbulence, which is characteristically caused by the R35S chamber. A larger amount of pre-mixture is formed at early pilot timing for R35S chamber. However, advanced pilot injection causes higher THC emission. This is because too long mixing period can over dilute the air-fuel mixture and misfire or incomplete combustion may occur.

Figure 4.11 shows combustion process with variable qp/qt, pilot injection timing θp= –15oATDC and θp= –60oATDC for each combustion chamber at high load engine condition. The injection timing is fixed at θi= 0oATDC. At any of these conditions, as qp/qt increase, rate of heat release produced from pilot combustion is also increased. Cylinder pressure and temperature at the moment before the main fuel injection start is higher. This can shorten the ignition delay period of main fuel injection and reduce maximum rate of heat release of premixed combustion. Increment of fuel delivery up to qp/qt=12% for R35S chamber with pilot injection timing θp= –15oATDC gives higher rate of heat release of pilot


Figure 4.9 Effect of pilot injection timing on combustion and emissions for STD chamber

-60 -40 -20 0200225250275300

00.51.01.52.0

0200400600800

0250500750

1000

050100150200

050100150

050100

45670

0.20.40.60.8

0100200300

0100200300

2468

-60 -40 -20 005101520

-60 -40 -20 0020004000

00.20.40.60.8

-60 -40 -20 0400500600700800

05101520

be g

/kW

h

S B

osch

TH

C p

pmC

STD 1800rpmP

e=0.7MPa

Pinj

=100MPaθ

i=TDC

qp/q

t=12%

θp oATDC θ

p oATDC

θp oATDCθ

p oATDC

NO

x pp

m

PA

RT

mg/

m3

SO

LID

mg/

m3

SO

Fm

g/m

3

w/o pilot

p max

MP

a

⊿p m

max

MP

a/de

g

⊿Q

m m

ax

J/de

g

STD 1800rpmP

e=0.7MPa

Pinj

=100MPaθ

i=TDC q

p/q

t=12%

Tex

h℃

⊿p p

max

MP

a/de

g

⊿Q

p m

ax

J/de

gQ

pilo

t/Qt

%

τ o C

A

w/o pilot

Opa

city

%

CO

ppm


Figure 4.10 Effect of pilot injection timing on combustion and emissions for R35S chamber

-60 -40 -20 0200225250275300

00.51.01.52.0

0200400600800

0250500750

1000

050100150200

050100150

050100

45670

0.20.40.60.8

0100200300

0100200300

2468

-60 -40 -20 005101520

-60 -40 -20 0020004000

00.20.40.60.8

-60 -40 -20 0400500600700800

05101520

be g

/kW

h

S B

osch

TH

C p

pmC

R35S 1800rpmP

e=0.7MPa

Pinj

=75MPaθ

i=TDC

θp oATDC θ

p oATDC

θp oATDCθ

p oATDC

NO

x pp

m

PA

RT

mg/

m3

SO

LID

mg/

m3

SO

Fm

g/m

3

w/o pilot

p max

MP

a

⊿p m

max

MP

a/de

g

⊿Q

m m

ax

J/de

g

R35S 1800rpmP

e=0.7MPa

Pinj

=75MPaθ

i=TDC

Tex

h℃

⊿p p

max

MP

a/de

g

⊿Q

p m

ax

J/de

gQ

pilo

t/Qt

%

τ o C

A

w/o pilot

Opa

city

%

CO

ppm


combustion and fluctuate the combustion pressure at TDC noticeably. This fluctuation is not seen for STD combustion chamber. Stratified rich combustion of R35S chamber can suppressed the initial combustion, however, due to high turbulence airflow, pilot combustion can become severe depend on pilot fuel delivery or pilot injection timing. However, promoted combustion can be seen for diffusion combustion. At both pilot injection timing θp= –15o ATDC and θp= –60o ATDC, rate of heat release for diffusion combustion are higher than STD chamber. This combustion characteristic is favorable with low initial combustion of pilot injection. Therefore, lenient pilot combustion can be obtained by using an advanced pilot injection timing θp=–60o

ATDC. Figure 4.12 shows exhaust emissions for Figure 4.11. In the

case of STD combustion chamber, the effect of pilot injection for NOx reduction is small. At θp= –15oATDC, NOx emission increase as pilot fuel delivery increase. This is due to higher cylinder pressure and temperature at TDC. On the other hand, rich combustion of R35S chamber gives no increment of NOx emission. Moreover, particulate emissions increase of using pilot injection for STD chamber at pilot injection timing θp= –15oATDC. The elimination of initial combustion causes the main combustion only comprising the diffusion combustion and increase the combustion duration. At pilot injection timing θp= –60oATDC, due to promoted air fuel mixture, increment of particulate emissions has been restrain. On the other hand, promoted diffusion combustion of R35S chamber due to high turbulence airflow has suppressed the increment of particulate emissions especially at pilot injection timing θp= –60oATDC which has lower particulate emissions than STD chamber. High turbulence of R35S chamber might promoted air fuel mixing more than STD chamber. 4.3.2 Effect at Medium Load Figure 4.13 shows combustion process at Pe=0.35MPa with pilot


injection of several pilot fuel delivery qp/qt. Here θi= TDC, θp=–15o ATDC and θp=–60oATDC. At θp=–15o ATDC , even at low pilot fuel delivery qp/qt=4%, initial combustion develops early and maximum rate of heat release reduced. At θp=–60oATDC condition, low pilot fuel delivery did not affect rate of heat release of main combustion. This is due to pilot fuel is diluted during long mixing period resulting in too low heat release of pilot combustion.

0

2

4

6

8

-30 -15 TDC 15 30

θp=-15oATDCSTD 1800rpm

Pe=0.7MPa

Pinj

=100MPaθ

i=TDC

θp=-60oATDC

p M

Pa

8

w/o pilotw/o pilot

8

12

qp/q

t=2%

12

⊿Q

p

θ oATDC

N.L.

qp/q

t=2%

0

50

100

150

200

⊿Q

J/

deg

-30 -15 TDC 15 30 45

0

2

4

6

8

-30 -15 TDC 15 30

θp=-15oATDCR35S 1800rpm

Pe=0.7MPa

Pinj

=75MPaθ

i=TDC

θp=-60oATDC

p M

Pa

8

w/o pilotw/o pilot

8

12 qp/q

t=2%

12

⊿Q

p

θ oATDC

N.L.

qp/q

t=2%

0

50

100

150

200

⊿Q

J/

deg

-30 -15 TDC 15 30 45

(a) R35S

(b) STD

Figure 4.11 Effect of pilot injection on combustion process


00 5 10200225250275300

00.51.01.52.0

0200400600800

0250500750

1000

050100150200

050100150

050100

00 5 10 1502000400005101520

be g

/kW

h

S B

osch

TH

C p

pmC

R35S 1800rpmP

e=0.7MPa

Pinj

=75MPaθ

i=TDC

qp/q

t % q

p/q

t %

NO

x pp

m

PA

RT

mg/

m3

θp=-15oATDC

θp=-60oATDC

SO

LID

mg/

m3

S

OF

mg/

m3

Opa

city

%

CO

ppm

00 5 10200225250275300

00.51.01.52.0

0200400600800

0250500750

1000

050100150200

050100150

050100

00 5 10 1502000400005101520

be g

/kW

h

θp=-15oATDC

θp=-60oATDC

S B

osch

TH

C p

pmC

STD 1800rpmP

e=0.7MPa

Pinj

=100MPaθ

i=TDC

qp/q

t % q

p/q

t %

NO

x pp

m

PA

RT

mg/

m3

SO

LID

mg/

m3

SO

Fm

g/m

3O

paci

ty%

CO

ppm

(a) R35S

(b) STD

Figure 4.12 Effect of pilot injection on emissions


Figure 4.14 shows combustion performance and exhaust emission characteristics with pilot injection against fuel delivery qp/qt at Pe=0.35MPa. Here, θi = TDC, θp = –15o ATDC and θp= –60o ATDC. As for combustion performance, at θp= –15o ATDC, τ are shorter, however, no significant change of τ due to increment of qp/qt. Here, as qp/qt increased, Qp max and ∆pp max increased while ∆Qm max and ∆pm max reduced. On the other hand, at θp= –60o

ATDC, as qp/qt increased τ is slightly shorten, and ∆Qm max and ∆pm max are slightly reduced. For exhaust emissions, at θp= –60o ATDC, increment of qp/qt cause higher THC concentration. This is because pilot fuel diluted to become lower than combustible limit due to intensified airflow of R35S chamber during ignition delay period, resulting misfired or incomplete combustion. For this reason, rate of fuel consumption, SOF and CO concentration are also increased.

Figure 4.13 Effect of pilot fuel delivery on combustion process under different pilot injection timings

0

2

4

6

8

w/o pilot

θθθθp=-15oATDCR35S

Pe=0.35MPa

θi=TDC

⊿Q

J/

deg

p

MP

a

22 15

qp/q

t=4%

⊿Q

p

-30 -15 TDC 15 30θ oATDC

N.L.

θθθθp=-60oATDC

0

50

100

150

200

w/o pilot

qp/q

t=4%

15

22

-15 TDC 15 30 45


4.3.3 Effect at Low Load Figure 4.15 shows combustion process at Pe=0.2MPa with injection timing with several fuel delivery qp/qt, θi=TDC and pilot Figure 4.14 Effect of pilot fuel delivery on combustion and emission

characteristics under different pilot injection timings

0 5 10 15 20200

250

300

0

1

0

500

1000

400

800

1200

0

50

100

0

50

0

50

4

6

8

0

0.4

100

200

300

0

100

200

300

4

8

0 5 10 15 20 250

10

20

0 5 10 15 20 250

2000

0

0.6

0 5 10 15 200

10

20

30

0

10

be g

/kW

h

R35SP

e=0.35MPa

θi=TDC

S B

osch

TH

C p

pmC

θp=-15oATDC

θp=-60oATDC

qp/q

t % q

p/q

t %

qp/q

t %q

p/q

t %

NO

x ppm

P

AR

Tm

g/m

3

SO

LID

mg/

m3

SO

Fm

g/m

3

θp=-15oATDC

p max

MP

a

θp=-60oATDC

⊿p m

max

MP

a/de

g

⊿Q

m m

ax

J/de

g

R35SP

e=0.35MPa

θi=TDC

⊿θ b

deg

⊿p p

max

MP

a/de

g

⊿Q

p m

ax

J/de

gQ

pilo

t/Qt %

τ o C

A

O

paci

ty%

CO

ppm


injection timing at θp= –15o ATDC and θp= –60o ATDC. At pilot injection θp= –15o ATDC, shortened ignition delay period and reduction of rate of heat release of main combustion can be seen even with low pilot fuel delivery. At injection pilot θp= –60o

ATDC with qp/qt=20%, slope of rate of heat release for main combustion increase gradually and timing of maximum rate of heat release move earlier towards TDC. Here, rate of heat release of main injection does not change. This is because at Pe=0.2MPa, total quantity of fuel delivery is low and even at qp/qt=20% pilot fuel delivery is still small in quantity, pilot fuel remain in combustion chamber as diluted mixture and then combust together with main injection fuel.

Figure 4.16 shows combustion performance and exhaust emission characteristics against qp/qt, at Pe=0.2MPa. Here, main injection timing θi=TDC and pilot injection timing at θp= –15o

ATDC and θp= –60o ATDC. At pilot injection θp= –15oATDC, ignition delay of main injection become shorter due to pilot injection and ∆Qm max and ∆pm max also reduced as qp/qt increased. At pilot injection timing θp= –15o ATDC, ∆Qp max and ∆pp max getting higher as qp/qt increased. However, at pilot injection

Figure 4.15 Effect of pilot fuel delivery on combustion process under different pilot injection timings

0

2

4

6

8

w/o pilot

θθθθp=-15oATDCR35S

Pe=0.2MPa

θi=TDC

⊿Q

J/

deg

p

MP

a

3320

qp/q

t=5%

⊿Q

p

-30 -15 TDC 15 30θ oATDC

N.L.

θθθθp=-60oATDC

0

50

100

150

200

w/o pilot

qp/q

t=5%20

-15 TDC 15 30 45


timing θp= –60o ATDC, increment of qp/qt does not give any change to Qpilot/Qt. This is because most of pilot fuel remains as

Figure 4.16 Effect of pilot fuel delivery on combustion and emission characteristics under different pilot injection timings

0 10 20 30250

300

350

0

1

0

500

1000

400

800

1200

0

50

100

0

50

0

50

4

6

8

0

0.4

100

200

300

0

100

200

300

4

8

0 10 20 30 400

10

20

0 10 20 30 400

2000

0

0.6

0 10 20 300

400

800

0

10

be g

/kW

h

R35SP

e=0.2MPa

θi=TDC

S B

osch

TH

C p

pmC

θp=-15oATDC

θp=-60oATDC

qp/q

t % q

p/q

t %

qp/q

t %q

p/q

t %

NO

x ppm

PA

RT

mg/

m3

SO

LID

mg/

m3

SO

Fm

g/m

3

θp=-15oATDC

p max

MP

a

θp=-60oATDC

⊿p m

max

MP

a/de

g

⊿Q

m m

ax

J/de

g

R35SP

e=0.2MPa

θi=TDC

Tex

h o C

⊿p p

max

MP

a/de

g

⊿Q

p m

ax

J/de

gQ

pilo

t/Qt %

τ o C

A

O

paci

ty%

CO

ppm


air-fuel mixture until the main combustion occurs. As for exhaust emissions, pilot injection have little effect on NOx emission concentration. At pilot injection timing θp= –60o

ATDC, rate of fuel consumption, THC and SOF concentration increase as qp/qt increased. 4.4 EFFECT OF HIGH TURBULENCE COMBUSTION

WITH EGR In this sub topic the investigation of the effect of EGR on engine performance and exhaust emission characteristics for R35S chamber in comparison with STD combustion chamber have been done. Figure 4.17 shows combustion process against crank angle at Pe=0.7MPa with EGR application both for R35S and STD chambers. Regarding the R35S chamber, the peak heat release rate at initial combustion stage is lower than the STD chamber because the R35S chamber has lower injection pressure. When EGR rate increases, heat release curve for both chambers show lenient slope just after the ignition. However, after lenient rise of heat release, the R35S chamber soon generates rapid heat release during initial combustion, which is independent of EGR rate. On the contrary, heat release curve of the STD chamber at initial burning shows lenient slope and peak heat release rate becomes higher with increasing EGR rate. It is noted that the R35S chamber gives higher heat release rate at the diffusion combustion stage compared with the STD chamber with or without EGR condition.

Figure 4.18 shows exhaust emission characteristics against EGR rate at Pe=0.7MPa for R35S and STD chambers. It can be seen that NOx concentration is reduced gradually as EGR rate increased. However, Smoke and SOLID emissions for STD chamber increase as EGR rate increased. On the other hand, R35S chamber can control the increment of smoke and SOLID emission as EGR rate increased.

From those results, it is known that NOx emission reduced


for both combustion chambers when EGR is applied. At Pe=0.7MPa with EGR application, PART emission for STD chamber getting worse, however, R35S chamber show special

Figure 4.17 Comparison of combustion process under EGR condition

0

2

4

6

8

w/o EGR

R35SPe=0.7MPa

θi=TDC

⊿Q

J/

deg

p

MP

a

23

EGR = 12%

⊿Q

p

-30 -15 TDC 15 30θ oATDC

N.L.

STD

0

50

100

150

200

w/o EGR

EGR = 12%

23

-15 TDC 15 30 45

Figure 4.18 Effect of high turbulence mixing on combustion and emission characteristics employing EGR

0 10 20200

250

300

0

2

0

500

1000

400

800

1200

0

200

400

0

200

0

200

0 10 20 300

4000

0

20

be g

/kW

hS

Bos

chT

HC

ppm

C

STD

Pe = 0.7MPa

θi = TDC

EGR % EGR %

NO

x ppm

PA

RT

mg/

m3

SO

LID

mg/

m3

SO

Fm

g/m

3

R35S

O

paci

ty%

CO

ppm


characteristics of controlling gradual increment of this emission. Figure 4.19 shows comparison of exhaust emissions for

all load engine conditions with variable EGR rate. NOx emission is reduced with EGR application for both combustion chambers. Rate of NOx emission reduction due to EGR application is higher at high load than low and middle load engine condition. Particulate and smoke emissions are almost zero for both combustion chambers with equivalence ratio less than 0.6. Particulate and smoke emissions deteriorate from equivalence ratio φ =0.6 for STD chamber and φ =0.7 for R35S chamber. However, compared with STD chamber, particulate emission deterioration for R35S chamber is small. At high load Pe=0.7MPa, EGR rate 20% is an operation limit due to too high emissions for STD chamber, while EGR rate 30% is an operation limit for R35S chamber. The increment of CO emission from φ =0.7 also indicates that promoted combustion due to high turbulence of R35S chamber at high load has great effect. Moreover, misfire or incomplete combustion at low load that might occur due to high EGR rate that will reduce oxygen concentration in the combustion cylinder have an effect to increase THC and rate of fuel consumption, however, the increment of THC and rate of fuel consumption is lower for R35S chamber. 4.5 MULTIPLE EFFECTS OF EGR, PILOT INJECTION

AND HIGH TURBULENCE COMBUSTION In this section, multiple effects on exhaust emissions of applying EGR and pilot injection with high turbulence combustion have been investigated.

Figure 4.20 shows exhaust emission characteristics against EGR rate at condition of mean effective pressure Pe=0.7MPa, injection timing θi= TDC, pilot injection timing θp= –15o ATDC and θp= –60o ATDC and rate of pilot fuel delivery qp/qt=12% for R35S chamber. Similarly with Figure 4.18, NOx concentration is reduced as EGR rate increase regardless of pilot injection.


(a) R35S

0.2 0.4 0.6 0.8200250300350400

012340

200400600800

0250500750

1000

0100200300400

0100200300

0100200

0.2 0.4 0.6 0.8 1.0020004000010203040

be g

/kW

h

S B

osch

TH

C p

pmC

40

3020

10

EGR=0%

0.20.35

0.5P

e=0.7MPa

R35S 1800rpmP

inj=75MPa

θi=TDC

φ φ

NO

x pp

m

PA

RT

mg/

m3

SO

LID

mg/

m3

S

OF

mg/

m3

Opa

city

%

CO

ppm

0.2 0.4 0.6 0.8200250300350400

012340

200400600800

0250500750

1000

0100200300400

0100200300

0100200

0.2 0.4 0.6 0.8 1.0020004000010203040

be g

/kW

h

S B

osch

TH

C p

pmC

0.640

3020

10

EGR=0%0.35

0.2

Pe=0.5MPa 0.7

STD 1800rpmP

inj=100MPa

θi=TDC

φ φ

NO

x pp

m

PA

RT

mg/

m3

SO

LID

mg/

m3

SO

Fm

g/m

3O

paci

ty%

CO

ppm

(b) STD

Figure 4.19 Emission characteristics against equivalence ratio under EGR


Smoke and SOLID emissions at θp= –15o ATDC become worse which has similar result as for only applying pilot injection. At θp= –60o ATDC, increment of Smoke and PART emissions are been controlled while NOx emission has been reduced. Rate of fuel consumption has less increment at this condition.

Figure 4.21 shows NOx and particulate trade-off relationship at Pe=0.7MPa, θi= 0o ATDC, qp/qt =11% with variable EGR rate and pilot injection timing or without pilot injection. NOx

emission is reduce as EGR rate increase, however, at pilot injection timing θp=–15oATDC, deterioration of particulate is severe with higher EGR rate. On the other hand, at pilot injection timing θp= –60o ATDC, particulate emission is lower than the condition of without pilot injection. Therefore, further reduction of NOx

emission can be obtained by increase the EGR rate at pilot injection timing θp=–60o ATDC.

Figure 4.20 Effect of pilot injection timing and EGR on emission

21 20 19 18 17 21 20 19 18 17 16 InO2 % InO2 %

00 10 20200225250275300

012340

200400600800

0250500750

1000

0100200300400

0100200300

0100200

00 10 20 30020004000010203040

be g

/kW

h

S B

osch

TH

C p

pmC

R35S 1800rpmP

e=0.7MPa

Pinj

=75MPaθ

i=TDC

EGR % EGR %

NO

x pp

m

w/o pilot

PA

RT

mg/

m3

SO

LID

mg/

m3

θp=-15oATDC

θp=-60oATDC

SO

Fm

g/m

3O

paci

ty%

CO

ppm


CONCLUSIONS The conclusions of the studies are as follow. i) Both combustion chambers has similar effect on NOx

reduction, however, R35S chamber can suppressed the deterioration of particulate and smoke emissions effectively. Therefore, high EGR rate application for R35S chamber has higher effect to reduce NOx emission.

ii) When high turbulence combustion chamber is applied together with significant advanced pilot injection, EGR can be set at high rate due to Smoke and PART emissions increment can be controlled. Furthermore, at the same time reduction of NOx emission can be obtained.

REFERENCES Cheolwoong, P., et al., “Effect of Multiple Injections in a HSDI

0 1 2 3 4 5 6 70

0.2

0.4

0.6

0.8

1.0

EGR

18%

6%

27%

22%

12%

EGR=0%

w/o pilot

θp=-15oATDC

θp=-60oATDC

R35S1800rpmP

e=0.7MPa

Pinj

=75MPaθ

i=TDC

qp/q

t=12%

PA

RT

g/

kWh

NOx g/kWh

Figure 4.21 NOx and particulate trade-off relationship at high load (R35S)


Diesel Engine Equipped With Common Rail Injection System”, SAE Paper 2004-01-0127, 2004.

Goda, E., et al., ”An Appraisal of High-Pressure Injection and High-Squish Combustion Chamber for Reduction of Diesel Particulate”, Trans. of JSAE, 34-3, pp49-54, 2003.

Kidoguchi, Y., et al., “Effect of High Squish Combustion Chamber on Simultaneous Reduction of NOx and Particulate from a Direct-Injection Diesel Engine”, SAE Paper 1999-01-1502, 1999.

Kidoguchi, Y., et al., “Experimental and Theoretical Optimization of Combustion Chamber and Fuel Distribution for the Low Emission Direct-Injection Diesel Engine”, ASME, Journal of Engineering for Gas Turbine and Power, Vol.125, p.351-357, 2003.

Kidoguchi, Y., et al., “DeNOx mechanism caused by thermal cracking hydrocarbons in the stratified rich zone during diesel combustion”, Int. J. Engine Res. Vol.6, 2005.

Mallmo, F., et al., “Analysis of Multiple Injection Strategies for the Reduction of Emissions, Noise and BSFC of a DI CR Small Displacement Non-Road Diesel Engine”, SAE Paper 2002-01-2672, 2002.

Yang, C., et al., “Effects of Fuel Properties on Combustion and Emission Characteristics of a Direct-Injectioin Diesel Engine”, JSAE trans., Vol.65, No.637, pp. 3203-3208, 1999. (in Japanese)

5

PERFORMANCE AND EMISSION EVALUATIONS OF A PROTOTYPE STEPPED-PISTON ENGINE USING

CARBURETOR AND DIRECT FUEL-INJECTION SYSTEMS

Azhar Abdul Aziz Zulkarnain Abdul Latiff Mohd Fawzi Mohd Ali

Mohd Farid Mohammad Said Mazlan Said

5.1 INTRODUCTION Two-stroke engines have been used for sometimes in automotive and stationary applications since early 20th century. The advantages of two-stroke engines are obvious, i.e., lighter, simpler and less expensive to manufacture. Technically, two-stroke engines have the potential to pack almost twice the power into the same space because there are twice as many power strokes per revolution. The combination of lightweight and twice the power gives two-stroke engines a great power-to-weight ratio compared to many four-stroke engine designs. However due to the short-circuiting process of the fuel before combustion, this has resulted in deterioration in overall performances especially poor combustion efficiency and high white smoke emission problem. Coupled with the improvement in the four-stroke engine technology, the former has overcome the latter in being the choice for mobile platform


applications. Due to high fuel cost and the need to explore the use of other fuel sources, notably gaseous fuels, a number of enthusiasts and engine developers have revisited the two-stroke engine design. Fuels such as hydrogen and methane are said to be ideal for use with the incorporation of the some new features (Goldsborough and Blaringan, 2003).

An engine design and development program was initiated at Universiti Teknologi Malaysia (UTM) in year 2003 to develop local R&D capabilities in small power-train engineering. The exercise evolved around the development of an air-cooled single cylinder of stepped-piston engine concept. The term “stepped-piston” refers to the conventional piston having compounded with a larger diameter section at the rear section of its geometry. The changes to the original design were made as the research group feels that there are rooms for improvements. In addition to this, the modifications will infuse other innovative scope of work from design to product testing activities (Hooper, 1985).This program, eventually leads to the incorporation of features, is expected to enhance performance of the prototype and subsequently exhaust emission. This is in anticipation of producing a working prototype for multiple applications namely stationary and automotive.

The gasoline stepped-piston engine is a relatively new design concept for small mobile power plants. It is an engine, operating on a two-stroke cycle but is infused with four-stroke engine features. It has a build-in supercharger mechanism (by virtue of the extended flange) that improves the scavenging process thus improve combustion efficiency. Due to these operating characteristics, the engine has all the attributes of a low emission, high-efficiency power plant that eliminates many of the major weaknesses associated with the Otto four-stroke engine and with modern two-stroke engines.

Performance And Emission Evaluations Of A Prototype 91

5.2 COMBUSTION GEOMETRY AND RECIPROCATING ELEMENTS

Bernard Hooper and his co-workers undertook the work on stepped-piston engines in the 1970s (Goldsborough and Blaringan, 2003). The unique feature lies in the piston geometry i.e. two concentric pistons combined into one solid piece.

Figures 5.1 (a) and 5.1(b) illustrate the engine design in detail. The double diameter pistons work inside a cylinder, which has two different diameters; the smaller providing a normal piston-ported 2-stroke cylinder (with combustion chamber) and the larger is the annular pumping chamber. As the piston moves downward, the intake charge (i.e. fresh air) is drawn into the annular space through the main reed valve. The reed valve will close automatically once the piston has reach bottom dead centre (BDC). On the up-swing motion of the piston the inducted air, that is now being trap, is force into the combustion chamber via a smaller set of reed valves, shown in Figure 5.2. Upon the completion of the combustion process, the exhaust gas will leave through the exhaust port in similar fashion to the current conventional two-stroke engine design.

(a) (b)

Figure 5.1 (a) Single-cylinder and (b) multi-cylinder arrangements

(Hooper, 1985)


Figure 5.2 illustrates the configuration made by the research group on the single cylinder stepped-piston design. Again, as mentioned earlier the inducted air will pass through a radial arrangement of intricate ports, leading to the main combustion chamber.

Figure 5.2 The cross-sectional area of the engine

The step or the skirting surface of the piston combines to act as a pump to induce fresh mixture during the expansion stroke. As for gasoline direct-injection (GDI) version, only air will be induced. During the compression-stroke the mixture will be compressed and pass through a set of reed valves and to the transfer ports into the main combustion chamber. During ascending the piston surface will cover the inlet ports and the trapped air (now in the annular section) will be guided to flow into a set of simple accumulators (refer Figure 5.2). In doing so it will damp the pressure surge and allows the air to be, expel completely into the combustion chamber.

By now, it can be said that some innovative changes have been made to the early design proposed by Hopper (1985). The changes are a three-port transfer system, shallow bowl-in-piston geometry, air-charge system and port timings. In the case of GDI the engine is incorporated with a direct fuel injector (DFI) taking position perpendicular to the surface of the piston. Due to geometrical constrain the spark plug is positioned 30o in relation to

Piston Step

Skirting surface

Annularsection

Reed

Accumulator Inlet and exhaust ports


the vertical axis. This is to provide an ideal configuration in providing a stratified-charge combustion feature, if lean-burn combustion strategy is to be use effectively in this engine. Details of the engine design; timing and major specifications are given in Figure 5.3, Figure 5.4 and Table 5.1.

Figure 5.3 (a) The stepped-piston engine cross-sectional area

Table 5.1 General specifications

Parameter Size/Feature Capacity (cc) 125

Bore (mm) 54.2

Stroke (mm) 54.2

Compression ratio 9.6

Ignition system Capacitive-Discharge

Fuel/scavenging systems Carburetor and Direct-fuel injection/

multi-port loop scavenge Cooling Air-cooled

Starting Starter motor

Max. power (kW@rpm) 7.6@7500

Calc max. Torque (Nm@rpm) 12.5@5000

Power-weight ratio (kW/kg) 0.46

INJECTOR

STEPPED PISTON

LUBRICANT METERING DEVICE


5.3 ENGINE SUB-SYSTEMS The improvised version of the engine features the incorporation of the following in-house technologies. 5.3.1 Lubricant-dispensing unit This is a built-in unit incorporated onto the engine and synchronized with the engine crankshaft. Its sole purpose is to dispense exact amount of lubricant, as and when, onto the surface of the piston skirt. In other words, the metering of the lubricant is made not for every engine cycle (as the norm for the conventional design), but on periodical basis. In doing so, will minimize the ‘carry-over’ effect of the lubricant into the combustion chamber thus drastically reduce the emission of heavy unburned hydrocarbons at the tailpipe. The engine’s electronic control unit (ECU) controls the timing for the dispensing.

Figure 5.4 Engine timing diagram

TDC

BDC

EO (98.99o)

EC (261.01o)

IO (132.01o)

IC (227.99o) Exhaust

Scavenging

Compression Expansion

Compression chamber cycle

Combustion chamber cycle

Induction (0o – 180o)

Compression (180o – 360o)

I (20o)

IC – Inlet valve closed IO – Inlet valve open


5.3.2 Breather Unit The sole purpose for the installation of this indigenously developed device is to trap lubricant from escaping the engine crankcase during cranking. For the reciprocating mechanism to operate effectively in the crankcase, air must be inducted into the engine lower chamber. Likewise, during expansion the inducted air must be purged out of the crankcase. The purging will also resulted in the lubricant to be expelled together with air, through the same passage in which the ventilate air is inducted. To avoid the depletion of the lubricant, a special breather was developed and installed. The retention efficiency of the breather unit is 99.3%. 5.3.3 Electronic Fuel Injector Controller (EFIC) The unit is designed as a compact unit and installed directly to the engine. It contains a printed circuit board (PCB) with electronic components. The CPU with onboard RAM and ROM will minimize the interface components. Pulse width modulation (PWM) for the fuel injector is done by a timer, which has been programmed to control the output pulse width. A multi terminal plug connects the EFIC to its sensors, injector, as well as to its power supply. The EFIC must withstand high temperature, humidity and physical stresses. The input signals for this EFIC are the engine speed, top dead center (TDC) marking and manifold absolute pressure (MAP). The MAP will be calibrated with the air-fuel-ratio (AFR) through experimental results. Performing thorough engine testing does the calibration process. The results then will be used as a look-up table, which will be programmed within the EFIC. Through sensors input, the best AFR will then be defined and correct mass of fuel will be injected to the combustion chamber. The amount of fuel injected will be control by PWM (output of the EFIC), which actuate the injector solenoid. Figure 5.5 illustrates the methodology or operational sequence in which


the actuation and the metering of fuel are done for the engine.

Figure 5.5 EFIC Flowchart

Figure 5.6 The prototype gasoline direct-injector 5.3.4 Fuel injection system Figure 5.6 shows the gasoline direct injector specifically developed for the engine. It consists of solenoid, casing, spring, needle, swirler, nozzle and lock. It features a hollow cone spray and capable of generating less than 20 micron droplet size. Also shown (Figure 5.7) is the dedicated direct fuel-injection system for the

Look-up Table

Identify Air-Fuel-Ratio

Pulse Width Modulation

Actuate Injector Solenoid

Sensors Signal (Speed, TDC and MAP)


prototype engine. Here the fuel is pressurized by an axial-type mechanical pump, which is driven via the engine crankshaft. A pressure relief valve is attached to the high-pressure fuel line to maintain the fuel pressure at 5.0 MPa while excessive fuel is returned back to gasoline tank. To reduce the fluctuation of the fuel pressure, a small fuel accumulator is also attached to the high-pressure fuel line. The fuel injector used is a prototype pressure-swirl injector with a static flow rate of 480 cc/min rated at 5.0 MPa. The spray produced by this type of injector is a hollow-cone and its nominal spray half-cone angle is 32°. The injector is driven by control module located in the Electronic Fuel Injector Controller (EFIC). Connected to the EFIC are i) speed sensor, ii) crank angle sensor, and iii) manifold absolute pressure (MAP). These sensors serve as input parameter for the EFIC to determine the correct injection timing and pulse width at any speeds.

Figure 5.7 The schematic of the direct fuel injection system

5.4 LABORATORY AND FIELD TRIALS A 19-inch rack exhaust gas analyzer system (TOCSIN IGD 300), complete with a 3-meter sampling probe was used for emissions measurements. The prototype engine was directly coupled to an

Axial-type mechanical high-

pressure fuel pump

Fuel from gasoline

tank

Fuel accumulator

Pressure relief valve

Fuel injector

High-pressure fuel line

EFIC

Engine speed sensor

Crank angle sensor

MAP sensor

+12 volts


eddy-current MAGTROL dynamometer equipped with a load controller. The engine, dynamometer and other auxiliary items are mounted on a seismic steel bed (2m x 4m) to cushion the excessive engine vibration emitted during the trial. Fuel consumption measurements were made using a precision Ono Sokki gravimetric fuel flow meter (the FP series, DF 0400/FM-2500). The BS 7698-1: 1993 standard for reciprocating engine test was used in the evaluation of engine performance parameters. The emission tests were conducted in parallel to the experimental work on the engine.

The engine with carbureted-fuel system was firstly put to test. Upon completion, the engine was removed with lubricant and spark plug replaced. It was then fitted with the newly developed direct-fuel injection system. All data acquired and processed are subsequently corrected to the standard ambient temperature and pressure.

The results of the engine full load tests and emissions are shown in Figures 5.8, 5.9, 5.10 and 5.11 respectively. Figure 5.8 clearly shows the differences in the profiles of specific fuel consumption parameter for the two engine versions. The results indicate the improvements made in fuel consumption, ranging from 5% to 7 %, when the engine is equipped with the GDI fuelling system.

The incorporation of the fuel injection system has also resulted in the improvement in the engine output, i.e., engine’s brake power, particularly from the mid-speed range onwards, as shown in Figure 5.9. It is worth mentioning that there is clear deviation between the two curves at the high end of the speed regime.

One of the problems with two-stroke engine is the rather excessive gaseous emission. As mentioned, earlier the short circuiting problem of two-stroke engine contributed to fuel losses as unburned fuel in the exhaust. With gasoline engines combustion the critical emission by-products are i) carbon monoxide (CO) and unburned hydrocarbons (uHCs). These parameters were observed based on a series of data gathered. With the incorporation of the


250

300

350

400

450

500

550

600

650

1000 2000 3000 4000 5000 6000 7000 8000 9000

Speed (rpm)

Sp

. Fu

el C

on

sum

pti

on

(g

m/k

Wh

)Carb. Version

GDI version

Figure 5.8 Engine brake power against speed at full load

0

2

4

6

8

10

1000 2000 3000 4000 5000 6000 7000 8000 9000

Speed (rpm)

Bra

ke P

ow

er (

kW)

Carb. Version

GDI version

Figure 5.9 Fuel consumptions against speed at full load


0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

1000 2000 3000 4000 5000 6000 7000 8000 9000

Speed (rpm)

Car

bon

mo

nox

ide

(gm

/sec

)

Carb. Version

GDI version

Figure 5.10 CO concentration against engine speed at full load

injector unit onto the engine and with precise timing and metering the reduction of these two emission parameters is seems obvious. Reduction of CO concentration ranges from 25% to 30% from low to high speed (Figure 5.10). Similarly, the reduction in un-burnt hydrocarbon emission (Figure 5.11) is also pronounced. The reduction of these two toxic components of the emission is attributed to i) high utilization of air for combustion, ii) improve combustion efficiency and iii) reduction in fuel loss.

The work by Orbital Corp Ltd and Mitsubishi Corp of Japan indicated that with the retrofitting of engine the fuel can be saved at a maximum of 12% (Sherman, 2008).

To ascertain the practicality of the newly developed engine, it was put to test in somewhat small-unmanned water surface craft (refer Figure 5.12). This is a craft, which is controlled remotely. It has radio coverage of 1 km in radius, whose sole purpose is for surveillance and water gauging. The craft was tested for sometimes


0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

1000 2000 3000 4000 5000 6000 7000 8000 9000

Speed (rpm)

un

bu

rnt

hyd

roca

rbo

n (

gm

/sec

) Carb. Version

GDI version

Figure 5.11 Unburned HC concentration against engine speed at

full load

Figure 5.12 Mounting of the prototype engine onto a remotely piloted water surface craft

The prototype engine in place


and the engine has proven to be reliable and durable for use in non-automotive sector such as this. The mode of the trial covers idle, full load and rapid throttling modes. However, the trials were only made on the carbureted version, not on the direct-injection version due to time and technical constrains. 5.5 CONCLUSIONS The following conclusions are hereby made:

i. A small-size gasoline engine (125cc) of stepped-piston

design has been successfully developed and test for both carbureted and fuel-injection versions.

ii. The incorporation of direct fuel-injection system has contributed towards the improvement of the engine performance and has shown to reduce HC and CO emissions respectively.

iii. The engine was subjected to a field trial and was found to perform well in a mobile platform.

As this is only the preliminary work on small engine programme, further refinements are required. The refinement will be towards the improvement of the power-to-weight ratio, and packaging of the sub-systems. This is will enable the engine attractive for many applications, namely in the non-automotive sectors.

REFERENCES Goldsborough, S.S and Blaringan, P.V. 2003. Optimizing the

Scavenging System for a Two-Stroke Cycle, Free Piston Engine for High Efficiency and Low Emissions: A Computational Approach. SAE Paper No. 2003-01-0001.


Hooper, B. 1985. Stepped Piston And Stepped Piston Engine. Patent Number 4,522,163, United State Patent, 11th. June 1985.

Sherman, D. 2008. 2009 Technology of the year: Direct Fuel Injection – latest news, features and review. Online Automobile Magazine.( http://www.automobilemag.com).

6 THE DEVELOPMENT OF GRAY CAST

IRON CYLINDER BLOCK FOR PASSENGER VEHICLE Raja Mazuir Raja Ahsan Shah

6.1 INTRODUCTION A gray cast iron cylinder block is the standard design for automotive applications. For diesel engine application, separate dry or wet liners which made from special wear-resistant materials are used to perform more reliable and durable product. While the cylinder blocks for truck engines continue to be manufactured in cast iron, aluminum passenger car blocks are becoming increasingly popular owing to their weight saving potential

This chapter is to study the process of the designing and development of gray cast iron cylinder block casting. The aims of this product is to reduce the product weight, increase the cold start conditions for emissions, improve the design passages of the oil breather system and trim down the rejection rate by improving the process quality. Due to the limited investment, the design of the casting product are being limited by engineering aspects such as jigs, fixtures and other related production processes.

In this work scope, it will emphasize on the casting product specifications and the problems or issues encountered during the development phase specifically on the breather cores and coolant jacket core. These issues have initially hampered the goals of the product, which then resolved through a proper study and adequate engineering tools to identify the root causes.

Research on Vehicle Technologies (2008)

106

6.2 CYLINDER BLOCK CASTING DEVELOPMENT Gravity casting process is selected due to its low investment cost. Because of the cost factor driven, the material for cylinder block has been decided to be JIS250 (gray cast iron) or equivalent mechanical properties. The properties of the JIS250 are given in Tables 6.1 and 6.2.

To produce cylinder block casting product, it requires nine sand cores and two patterns, namely:

i. Coolant Jacket Core ii. Intake Oil Core iii. Exhaust Oil Core iv. Front Core v. Rear Core vi. Case No 1 Core vii. Case No 2 Core viii. Case No 3 Core ix. Case No 4 Core x. Drag Pattern xi. Cope Pattern

The modeling of the cylinder block cores and patterns has

taken into account the design and manufacturing parameters including the cylinder block machining line. The platform used to produce these models is CATIA 4.2.4, a commercial Computer Aided Engineering (CAE) software package, that is capable to perform fast track tasks, i.e., designs and manufacturing analysis. Figure 6.1 shows a complete model of a cylinder block casting.

The Failure Method Engineering Analysis (FMEA) method with casting aspects has been used as an approach to identify the potential issues and concerns during production process. Further development of casting process was established to reduce cost and lead-time as well as improving the quality.

The quality of the cylinder blocks casting product is to be controlled during core making and casting processes, where no unacceptable defects should be appeared on casting surface. Only

The Development of Gray Cast iron Cylinder Block for Passenger 107

small amount of sand can be allowed after casting process to prevent any damage to the engine components, i.e., coolant pump, oil pump and etc. To avoid any excessive corrosion on the cylinder block, external surface will be painted in special synthetic resin coating.

Figure 6.1 Cylinder block 3-D casting model Catia V4.2.4

6.3 CASTING SPECIFICATIONS New casting specifications have been developed to ensure that the product can meet the quality and other desired parameters such as functionality, durability and reliability. All of these specifications are controlled in cylinder block casting specification documents and to be used as a tool at production line to check into detail of the casting product for conformity to the standards.

The casting quality shall be homogeneous and free from scale, visible blowholes and loose or adhering sand. The surface roughness of the casting shall not exceed 200Rmax and the surface of the casting shall have no raised or coarse substances such as burnt-on sand, slag and etc. The permissible sand quantity in casting is shown in Table 6.1.


108

Due to high combustion pressure and loads, the block casting shall not be repaired by metal insert, welding, soldering or other methods. The tolerances for the casting product are specified to prevent any interference with others engine components and jigs/fixtures of the machining line, where for general wall thickness is set at t0.1

5.0+− where t is defined by Computer Aided

Drafting (CAD) model and around oil drains is t5.10.1

+− . For other

general tolerances on cast surface relative to datum, the details are given in Table 6.2.

Table 6.1 Amount of permissible sand in casting product

Remaining Sand (grams)

Area Loose Sand Before Painting

Lose Sand Before

Machining

Finished Machining

Coolant Jacket 2.50 2.50 0.50 max Oil Drains 2.50 2.50 0.01 max

To prevent the casting product from corrosion and to prolong its reliability against the extreme environment, a painting specifications derived from JIS have been adopted for this product as listed in Table 6.3.

For the material, a low alloy gray cast iron is chosen due to the rationale explained earlier in this chapter. The specifications are taken from JIS FC250 and the chemical and mechanical properties are shown in Table 6.4 and Table 6.5 respectively.

Table 6.2 General tolerance for casting product

Dimensions

(mm) Single Mould

(mm) Multiple Moulds

(mm) 100 or less ±1.0 ±1.5 101-200 ±1.5 ±2.0 201-400 ±2.0 ±3.0 401-800 ±3.0 ±4.0


Table 6.3 Painting specifications for cylinder block casting

Coating Special synthesis Resin Coating (Nippon Yushi Chassis Black M or equilvalent

Color Black Diluents Thinner for Chassis Black M Painting process Spray coating Drying process Radiative heat baking for 12 mins or

more Quality of paint & corrosion resistance(1)

FPO-1 for 8 hours or more

(1)Red rust shall be 10% or less. Film thickness 10µm or more. No cross cut at salt spray test

Table 6.4 Chemical composition of JIS FC250

Chemical Composition (%)

C Si Mn P S Cr Cu

3.10-3.50

1.80-2.40

0.50-1.00

0.15-max

0.15-max

0.15-0.35

0.30-0.55

Table 6.5 Mechanical properties of FC250

Area Brinell Hardness (HB) Casting 179-269

Cylinder Bore Sectional Area Min 187 6.4 DESIGN AND PROCESS ISSUES During the experimental works, two major design and process issues have been encountered which affect the overall performance of the product. The design of the oil breather system has been compromised as the result of the tight packaging around the main oil gallery. The consequence of this design has led a small cross section area that weakens the strength of the core. The rejection rate during core making process is high due to the process handling


110

as well as casting process which will increase the pressure in the crankcase system because of the oil passages blockage. The remaining sand also damages the moving components such as oil pump, pistons and valves.

It was also noticeable that the same issues happened at coolant jacket core. The root cause for this concern is the design of the tie-plate that is used to strengthen the coolant jacket core. The severity of the concern is critically high due to the damages it can cause to the overall powertrain system, i.e., total damage due to engine overheating. An engineering tool and repeatable experimental works have been conducted to resolve this issue. 6.4.1 Intake Oil Core Breakage The area around the oil gallery for intake oil core has made the core to be in thin section due to limitation of packaging and hence reducing the strength of the bonding sand. FMEA has been carried out to determine the possibilities of any core breakage during manufacturing process.

Few methods were introduced to reduce the risks and to provide a countermeasure to this issue. i. Cold Box core of original design in zircon sand ii. Shell sand cores of increased width in silica sand iii. Cold box cores of original design in zircon sand and steel

insert iv. Cold box cores of increased width in zircon sand

Based on these four solutions, test procedures and equipment were set up to confirm and validate the designs. The result will be used as a baseline for design team to continue with the design modifications. 6.4.1.1 Test Rig A test rig as depicted in Figure 6.2 was designed and built to break


the samples of cores transversely under direct dead weight load. The core supports are topped with 6 mm diameter rod, to give smooth bearing surface and set at 230 mm.

The height of the support differs by 27 mm such that the central necked portion of the core holds a stirrup, hook and plastic container. Lead shot is fed into the container from a reservoir through a ball type valve. When the test piece breaks, the valve is shut off promptly by hand. The breaking load is the total weight of the pin, stirrup and hook, plus the plastic container with its lead shot. Figure 6.3 shows the test rig setting.

6.4.1.2 Trial Cores The cold box process using zircon sand made the original cores. A second core box was manufactured in metal with the necked area increased in width for additional strength. This box was used to make cold box cores in zircon sand and shell process cores (hot box) in silica sand.

A further trial was based on the incorporation of a steel insert into necked area of the original core box prior to blowing the core to impact the strength to the core.

Figure 6.2 Oil breather core test rig


112

Figure 6.3 Test rig setting 6.4.1.3 Test Methods Ten cores of each of the four types were divided into a random order and the four types are as follows: i. Cold box cores original design in zircon sand (core weight

40 gm). ii. Shell sand cores of increased width in silica sand (core

weight 250 gm) iii. Cold box cores of original design in zircon sand and steel

insert (core weight 415 gm) iv. Cold box cores of increased width in zircon sand (core

weight 445 gm)

6.4.1.4 Test Results Based on calculation of loading cases under bending as shown in Figure 6.4.

F

l

bFA = and )1.6(F

l

aFB =


)2.6(3

22

aIE

F

l

bas

×××=

Where,

a = 114 mm b = 107 Ia = 25.17 x 106 mm4

The tensile strength of the core at the critical area is given by equation

)3.6(A

F=σ

where F is the load of the bucket and A is the sectional area at the point of load.

The tensile stress is dependent on the type of sand used to make the core. The full set of results are tabulated and graphed in Table 6.6. The original design cores gave an average transverse strength of 2,895 grams and increasing the width of the necked are raised this to 4,727 grams. The shell cores, also of increased width, gave similarly good results at an average of 4,064 grams.

Figure 6.4 Oil breather cores test load

F, Load (kg)

Fixed Support, FA

Fixed Support, FB

s

a

l

b


114

The introduction of steel inserts into the core box

interrupted the smooth flow of the sand into the necked area, resulting in very low strengths from the poorly compacted sand. The average strength using the steel inserts was reduced to only 1,401 grams

Hence from the table, the lowest tensile strength (original design with insert) is given by,

2/7.5

261

1401

mmgm

A

F

=

=

=σ

where, area is 261 mm2 for original design and for new design with widen cross sectional area

2/1.16

294

4727

mmgm

A

F

=

=

=σ

The cores for production casting will be made in the hot

box. The sand to be used for these trials was based on Kings Lynn AFS 100 Grade sand with 2.9% resin.

The sand for production will have similar content, but with a grading of 85 – 92 AFS. Whilst there may not be significant differences between these two sands, the only true comparison will be from practical trials.

From the result of the trial tabulated in Table 6.6, it can be concluded that to minimize the risks and costs, Zircon sand is necessity due to its strength (based on the tensile strength calculation) and grain structure. Rheotec coating to be applied to weak area, mainly on intake oil core to strengthen the core.


Table 6.6 Test result of various core design under loads

Core Design

Original Shell (wide) Insert Wide

Core

Load (gm) 1 2980 4580 853 4425 2 2250 2950 1385 4535 3 3175 3185 1275 5095 4 3085 4140 1050 4540 5 4115 4325 1675 5160 6 2620 3700 1135 4035 7 2890 3825 1925 3705 8 2540 3810 2130 4905 9 2815 4585 1595 5305 10 2480 5550 985 6565

Mean 2895 4046 1401 4727

6.5 COOLANT JACKET BURNT-ON SAND The burnt-on sand appears to happen during production trial. Two types of burnt-on sand observed are: i. Loose metal built / bridging from Coolant Jacket outer wall to

Tie Plate, blocking the Coolant flow through 2 mm channel. ii. Occurrence rate 100%, localize on Drag (Bottom pattern) side

only. Possible root cause is that the hot metal penetrates into coolant jacket core (2 mm thickness area as shown in Figure 6.5a and Figure 6.5b) during pouring process. This is due to insufficient sand compaction around the edge of tie-plate that causes the core to break when molten metal flows around the coolant jacket core. The molten metal then will start to flow into the crack area and form a fin as well as burnt-on sand as shown in Figure 6.6. This blockage may increase the coolant temperature especially during high engine revolution due to insufficient heat transfer from the combustion gas to coolant and then to radiator as a result of


116

turbulence and small velocity of flow around tie-plate area. The sand may also become loose due to coolant impact. This loose sand will carry into the coolant pump and can damage the impeller and the seal. As a result, the coolant pump will leak or not functioning that will increase the engine coolant temperature. Two methods of solution were identified to resolve the issue after considering the time and costs factors. They are: i. To apply core coating to strengthen the core. Core coating will

be applied on weak area only. ii. To use a new design of tie-plate with “0” gap (Figure 6.7 and

Figure 6.8 show the differences).

Figure 6.5a Cylinder block cross-section area

Figure 6.5b 2 mm Coolant channel blockage due to burn-on sand

Figure 6.5b

2


Figure 6.6 Location of Sand Burnt-on defect

Figure 6.7 Old tie-plate design


118

Figure 6.8 New tie-plate design Table 6.7 Results for various design of tie-plate and core coating

Trial No Description Result

MD1 Increase the viscosity of core coating (Rheotec 366M)

Not Good

MD8 Use new type of core coating (Rheotec 460)

Not Good

ME29 Use new tie plate design Good MG17 Use new tie plate design Good

Series of trial have been conducted to find the best solution to overcome the problem. The trial was conducted at OEM Casting Plant using TSC sand and two types of core coating, Rheotec 366M and Rheotec 460 respectively. Tie-plate design was fabricated by using wire cut process based on new design. The results of the trials are shown in Table 6.7.

Based on the result, the viscosity of core coating will not increase the strength of the sand core. It’s only improving the casting surface appearance and it was very obvious to see the burnt-on sand after both experimental trials. It can be concluded that core coating will not strengthen the core design and bonding structure of the sand.

The next results tell that the new tie-plate design has improved the defect of the sand core, hence eliminates the sand burn in problem. Two trials, ME29 and MG17, gave good results with a minimal burnt-sand that comply with Table 6.1 requirements. The other thing that can be observed is that the new tie-plate design has improved the coolant flow towards bore bridging area; hence increase the heat transfer at 30% of the stroke (Top Dead Centre) during combustion process to minimize the bore distortion.

6.6 CONCLUSION In this chapter, two major concerns have been studied and


solutions have been proposed to resolve them. From the FMEA and experimental methods, the breather core failure is due to the selection of sand and the design. By increasing the cross section area at the weak section, changing the sand from RCS to Zircon and applying the Rheotec coating at weak area has improved the strength of the breather core.

For the coolant jacket core, the burnt-on sand issue is easily resolved by changing the design of the tie-plate. The design also can improve the coolant flow towards bore bridging area; hence increase the heat transfer at 30% of the stroke (Top Deck Centre) during combustion process to minimize the bore distortion. This is yet to be proven during the 400 hours durability, thermal shock and cold start tests. REFERENCES Ataides, R. 2008. Numerical Analysis of Flow at Water Jacket of

an Internal Combustion Engine. SAE Technical Paper No. 2008-01-0393.

Canter, N. 2005. Cleaner Combustion Through Better Air/Oil Separation. Tribology and Lubrication Technology, 61 (12): 16-17.

Gallo, S. and Mus, C. 2000. Current Quality Needs for Casting in Automotive. Proceedings of the Merton C. Flemings Symposium on Solidification and Materials Processing, pp. 373-378.

Gerhard, W. 1995. Casting Production Expands at VAW. Foundry International, 18 (4): 1-5

Geus, D. and Stiebler, M. 2008. Industrial Application of Advanced Measuring and Evaluation Methods for Cylinder Liners of Engine Blocks. Journal Measurement Science and Technology, 19 (6): 064004 (1-8).

Grout, J.R. 2002. An Exercise in Mistake-Proofing and Failure Mode Effects Analysis. Proceedings - Annual Meeting of the Decision Sciences Institute, pp. 871-876.


120

Kajikawa, H., Nomura, S., Yasuda, H. and Mihara, M. 1993. Development of Light Weight Cylinder Block. SAE Technical Paper No. 938026

Kessler, M.P., Kruger, M., Ataides, R., Siqueire, C.D.L.R., Argachoy, C. and Mendes, A.S. 2007. Numerical Analysis of Flow at Water Jacket of an Internal Combustion Engine. SAE Technical Paper No. 2007-01-2711.

Noh, Y.K., Lee, B.H and Park, K.S. 1991. Fabrication of Automotive Components by Applying Cast Iron Chips. SAE Technical Paper No. 912538.

Schwaderlapp, M., Wagner, T., Hinz, R. and Peters, B. 1997. Lightweight Design of Future Engine Families SAE Technical Paper No. 97A086.

Semp, B.W., Pathan, A. and Dessert, P.E. 2006. The Role of Automated FMEA in Automotive Reliability Improvement. SAE Technical Paper No. 2006-01-1619.

Smetan, H. 2007. Forward Looking Light Engine Construction in The Stress Field of Casting Processes and Materials. VDI Berichte (1949), pp. 97-129.

Stamatis, D.H. 1996. FMEA and the Qs-9000 Requirement. SAE Technical Paper No. 961265.

Tomio Okamura, Takuro Sato, Sadao Kitazawa, Kazuo Nishidate 2000. Thin wall and lightweight cylinder block production technology. SAE Technical Paper No. 2000-05-0067.

Wissussek, D. and Schafer, R. 2002. Extending the Oil Change Intervals [of Automobiles] by Constructive/Conceptual Measures to the Engine. Tribologie und Schmierungstechnik, 49 (2): 10-14.

7

IDENTIFICATION OF VEHICLE SUSPENSION SYSTEM USING

PARTICLE SWARM OPTIMISATION WITH NEURAL NETWORK

Musa Mailah Gigih Priyandoko

Hishamuddin Jamaluddin

7.1 INTRODUCTION It is usual that acquiring the mathematical model of a vehicle or its subsystems is a prerequisite to designing a high-performance controller. In the past, most vehicle models have been constructed analytically in the form of dynamic equations. A conventional vehicle model, however, suffers from structural complexity, long development time, and the difficulty of modelling the highly nonlinear terms and the measurement noise. In order to control a real vehicle beyond mere simulation, it is indispensable to model the vehicle quickly, easily and as accurate as possible (Yim and Oh, 2004). Joo et al. (2000) developed a comprehensive nonlinear model of a vehicle suspension system which is derived using the standard kinematics and kinetics that takes into consideration a number of factors that are neglected in most existing models.

Typical control strategies rely on linear and time-invariant models. Buckner et al. (2000) used neural network that continually learns and estimates the nonlinear parameter variations of a quarter-car suspension model. This estimation algorithm becomes the foundation for an Intelligent Feedback Linearization controller


for active vehicle suspensions. Billings et al. (1992) demonstrated that NN could be used successfully for the identification and control of non-linear dynamical systems. The most advantageous and distinguishing feature of NN is their ability to learn. The network in the adaptive mode abstracts and generalizes the function character in the process of learning from training patterns. The learning algorithm is an optimization method capable of finding weight coefficients and thresholds (learning rates) for a given neural network and a training set. The learning algorithm that is used most frequently is the back propagation (BP) method. Although BP training has proved to be efficient in many applications, its convergence tends to be slow, and yields to suboptimal solutions (Billings et al., 1992; Narendra, 1996). To counter this problem, a particle swarm optimization (PSO) technique is proposed and incorporated into the NN system. Generally, the PSO is characterized as a simple heuristic of a well balanced mechanism with flexibility to enhance and adapt to both global and local exploration abilities. It is a stochastic search technique with reduced memory requirement, computationally effective and easier to implement (Kennedy and Eberhart, 1995; Parsopoulos and Vrahatis, 2002; Van den Bergh, 1999).

The chapter is organised as follows: Section 7.2 describes the dynamics of the vehicle suspension system while section 7.3 presents the structure of the NNs using BP learning algorithm. Section 7.4 describes the Particle Swarm Optimisation technique. The results of the simulation study are discussed and presented in section 7.5. Finally, the chapter is concluded in section 7.6. 7.2 QUARTER CAR VEHICLE SUSPENSION A standard assumption in the design and analysis of controllers for vehicle suspension system is that the vertical vehicle dynamics can be modeled using four independent quarter car suspension models. Quarter car models are simple but yet capture many important

Identification Of Vehicle Suspension System Using Particle Swarm 123

characteristics of the full model. Figure 7.1 depicts a typical laboratory scale two degree-of-freedom quarter car model test rig of a passive suspension system, in which the single wheel and axle are connected to the car body through a passive spring-damper combination. Figure 7.2 shows the representation of the same system that can be used for the computation of the mathematical model based on a number of realistic assumptions.

Figure 7.1 A quarter car suspension system test rig


Figure 7.2 A representation of a quarter car vehicle suspension The parameters shown in Figure 7.2 are defined as follows:

ms : sprung mass mu : unsprung mass bs : damping coefficient ks : spring stiffness coefficient kt : tyre stiffness coefficient zs : displacement of the car body (sprung mass) zu : displacement of wheel (unsprung mass) zr : displacement of road zs–zu : suspension deflection zu – zr, : tyre deflection

sz& : velocity of sprung mass

uz& : velocity of unsprung mass

sz&& : acceleration of sprung mass

uz&& : acceleration of unsprung mass


7.2 NEURAL NETWORK NN itself is basically a model structure and contains an algorithm for fitting the model to some given data. The network approach to modelling a plant uses a generic nonlinearity and allows all the parameters to be adjusted. In this way it can deal with a wide range of nonlinearities. Learning is the procedure of training the NN to represent the dynamics of a plant. The NN is placed in parallel with the plant and the error between the output of the system and the network output, the prediction error, is used as the training signal. NN have a potential for intelligent control systems because they can learn and adapt, approximate nonlinear functions, suited for parallel and distributed processing, and they naturally model multivariable systems. The advantageous and distinguishing feature of NN is their ability to learn. The learning algorithm is an optimization method capable of finding the weight coefficients and learning rate for a given NN and a training set.

Figure 7.3 The architecture of the proposed NN model


This algorithm is based on minimizing the error of the NN output compared to the required output. The required function is specified by the training set. The error of network E relative to the training set is defined as the sum of the partial errors of network Ek

relative to the individual training patterns and depends on network configuration w (Wilamowski, 2003; Wilamowski et al., 1999):

( )∑ ∑= =

−==p

k

p

kkkk dOee

1 1

2

2

1 (7.1)

where p is number of available patterns, ek is partial network error, Ok is output of neural networks, dk is teach data or desired output. Updating the weights of each layer using BP method in time t > 0 is calculated in order to minimize the error as follows (Wilamowski, 2003; Wilamowski et al., 1999):

( ))2()1()1(

)1(

)1()1()(

−−−+∂

−∂=−∆

−∆+−=

twtww

tEtw

twtwtw

jkjkjk

jk

jkjkjk

βα (7.2)

where 0 < α < 1 is the learning rate, β is the momentum. The speed of training is dependent on the set constant α. If a low value is set, the network weights react very slowly. On the contrary, if a high value is set, the algorithm fails. Therefore, the parameter α is typically obtained experimentally. The structure of the multilayer NN used in the study consists of the input, output and hidden layers. The input layer has three inputs represented by the road profile, sprung mass acceleration and suspension deflection. The output layer has two outputs, namely the sprung mass acceleration and suspension deflection. Every output neuron uses a linear activation function. Hidden layer have three neurons and each neuron uses sigmoid bipolar activation. Figure 7.3 shows the


architecture of the proposed NN model. 7.4 PARTICLE SWARM OPTIMIZATION The PSO idea was originally introduced by Kennedy and Eberhart in 1995 as a technique through individual improvement plus population cooperation and competition, which is based on the simulation of simplified social model, such as bird flocking, fish schooling and the swarm theory (Kennedy and Eberhart, 1995). Nowadays, PSO has gained much attention and wide applications in various fields (Lu et al., 2002; Guo et al., 2006).

The basic PSO algorithm consists of three steps, namely, generating particles’ positions and velocities, velocity update, and position update. Here, a particle refers to a point in the design space that changes its position from one move (iteration) to another based on velocity updates. First, the positions, i

kx , and velocities, ikv , of the initial swarm of particles are randomly generated using

upper and lower bounds on the design variables values, minx and

maxx , as expressed in Equations (7.3) and (7.4). The positions and velocities are given in a vector format with the superscript and subscript denoting the i th particle at time k. In Equations (7.3) and (7.4), rand is a uniformly distributed random variable that can take any value between 0 and 1.

)(0 minmaxmini xxrandxx −+= (7.3)

t

xxrandx

time

positionv minmaxmini

∆−+

==)(

0 (7.4)

The second step is to update the velocities of all particles at

time k+1 using the particles objective or fitness values which are functions of the particles current positions in the design space at time k. The fitness function value of a particle determines which


particle has the best global value in the current swarm, gkp , and also

determines the best position pbest of each particle over time, pi, i.e. in current and all previous group moves gbest. The velocity update formula uses these two pieces of information for each particle in the swarm along with the effect of current motion,i

kv , to provide a

search direction, ikv 1+ , for the next iteration. The velocity update

formula includes some random parameters, represented by the uniformly distributed variables, rand, to ensure good coverage of the design space and avoid entrapment in local optima. The three values that effect the new search direction, namely, current motion, particle own memory, and swarm influence, are incorporated via a summation approach as expressed in Equation (7.5) with three weight factors, namely, inertia factor, w, self confidence factor, c1, and swarm confidence factor, c2.

( ) ( )t

xprandc

t

xprandcwvv

ik

gk

ik

iik

ik ∆

−+

∆−

+=+ 211 (7.5)

The position of each particle is updated using its velocity

vector given by:

tvxx ik

ik

ik ∆+= ++ 11 (7.6)

7.5 RESULTS AND DISCUSSION The architecture of the suspension dynamics identification using NN technique is illustrated in Figure 7.4. Data for the training of the NN were extracted from the physical experimental rig with sprung mass weighing approximately 150 kg, unsprung mass 35 kg and tyre pressure equals to 20 psi. Sprung mass (vertical body) acceleration, suspension deflection and tyre deflection were considered as the output variables.


Figure 7.4 The architecture of the suspension dynamics modelling

The input variable comes in the form of the road profile. Physical sensors were incorporated to obtain the required input/output signals for identification purpose and they were connected to a PC-based data acquisition system (DAS).

Accelerometer was installed at the sprung mass of the vehicle suspension to measure body acceleration, a linear variable differential transformers (LVDT) was placed in between the sprung mass and unsprung mass to measure suspension deflection and another LVDT is attached between the unsprung mass and input disturbance (road profile) to measure the vertical displacement of road profile. Two types of road profiles were used in the form of approximate sinusoidal and pulse waves as depicted in Figures 7.5 and 7.6 respectively. Both the road profiles were generated by a specially designed pneumatic system controlled by a programmable logic controller (PLC).

The training algorithm for the proposed NN in this study is BP with PSO. It is usual that the NN parameters related to the


Figure 7.5 Sinusoidal wave road profile

Figure 7.6 Pulse wave road profile


weights, biases and learning rates (thresholds) of the BP algorithm are randomly initialized. The pseudo code of the training procedure is given as follows:

Begin PSO For each particle Initialize particle (v0 and p0) End Do For each particle Calculate fitness value If fitness better than pbest update pbest End Determine gbest from all particles For each particle Update velocity to formula (5) Update position to formula (6) End While maximum iterations or minimum error criteria is not attained Begin Neural Network Initialize weights (wi) and learning rateθ Do Input xi(t) with desired output d(t) Calculate error to formula (1) Adapt weights to formula (2) While not done

Moreover, the parameters of the NN were determined by

using PSO method in an off-line manner after a number of trial runs. Subsequently, the setting parameters of the PSO algorithm were obtained as follows: number of particles = 50, dimensions = 20, c1 = 1.25, c2 = 1.25 and w = 0.35. Results obtained from the simulation were shown in Figures 7.7 and 7.8 related to the identified sprung mass acceleration and suspension deflection respectively for a sinusoidal wave road profile.

Figures 7.9 and 7.10 are the identified sprung mass acceleration and suspension deflection respectively for a given pulse wave road profile.


Figure 7.7 Sprung mass acceleration of sinusoidal wave road profile

Figure 7.8 Suspension deflection of sinusoidal wave road profile


Figure 7.9 Sprung mass acceleration of square wave road profile

Figure 7.10 Suspension deflection of square wave road profile

In all the figures, the parameters to be identified are the actual acceleration and suspension deflection curves of the quarter car represented by the blue solid lines. The black dotted lines show


the identified parameters using NN with BP algorithm, while the red solid lines are those identified by the NN with PSO scheme. It is very obvious that the latter scheme performs better than the NN scheme without PSO learning algorithm since they almost replicate the actual responses. 7.6 CONCLUSION An alternative approach to modelling a passive vehicle suspension system using NN with PSO training method has been presented and successfully applied. By using the experimental quarter car test rig data, the nonlinear characteristics of the vehicle suspension system can be captured without having to resort to its dynamic model. The NN model responses and the actual test rig outputs are almost identical, implying that the NN model was able to capture the real vehicle dynamic characteristics. Further rigorous investigation should be carried out to evaluate the proposed model performance compared with other methods. The results of this study can also be used as a basis to design more complex suspension control system involving intelligent technique. REFERENCES Billings, S.A., Jamaluddin, H.B. and Chen, S. 1992. Properties of

Neural Network with Applications to Modelling Non-linear Dynamic Systems. International Journal of Control, 55(1): 193-224.

Buckner, G.D., Schuetze, K.T. and Beno, J.H. 2000. Active Vehicle Suspension Control Using Intelligent Feedback Linearization, Proceedings of the American Control Conference, Chicago, Illinois, pp. 4014-4018.

Guo, Q.J., Yu, H.B. and Xu, A.D. 2006. A Hybrid PSO-GD Based Intelligent Method for Machine Diagnosis. Digital Signal


Processing, 16(4): 402-418. Joo, D.S., Al-Holou, N., Weaver, J.M., Lahdhiri, T. and Al-

Abbas, F. 2000. Nonlinear Modelling of Vehicle Suspension System, Proceedings of the 2000 American Control Conference, Vol. 1, Issue 6, pp.115-119.

Kennedy, J. and Eberhart, R.C. 1995. Particle Swarm Optimisation. Proceedings of IEEE International Conference on Neural Networks, Vol. IV, pp. 1942-1948.

Lu, W.Z., Fan, H.Y., Leung, A.Y. T. and Wong, J.C.K. 2002. Analysis of Pollutant Levels in Central Hong Kong Applying Neural Network Method with Particle Swarm Optimization, Environmental Monitoring and Assessment, 79(3): 217–230.

Narendra, K.S. 1996. Neural Networks for Control: Theory and Practice, Proceedings of IEEE, 84(10): 1385–406.

Parsopoulos, K.E. and Vrahatis, M.N. 2002. Recent Approaches to Global Optimization Problems Through Particle Swarm Optimization, Natural Computing, 1(2-3): 235–306.

Van den Bergh, F. 1999. Particle Swarm Weight Initialization in Multi-layer Perceptron Artificial Neural Networks. Proceedings of ICAI, Durban, South Africa, pp. 41-45.

Wilamowski, B.M. 2003. Neural Network Architectures and Learning. Proceedings of IEEE International Conference on Industrial Technology, Vol. 1.

Wilamowski, B.M., Chen, Y. and Malinowski, A. 1999. Efficient Algorithm for Training Neural Networks with One Hidden Layer. Proceedings of the International Joint Conference on Neural Network (IJCNN), Vol. 3, pp. 1725-1728.

Yim, Y.U. and Oh, S.Y. 2004. Modelling of Vehicle Dynamics From Real Vehicle Measurements Using a Neural Network With Two-Stage Hybrid Learning for Accurate Long-Term Prediction. IEEE Transactions. on Vehicular Technology, 53(4): 1076-1084.

136 Index

INDEX Research On Vehicle Technologies

Active Control, 16,18,42 active force control, 41,43,45 Astrom-Hagglund Method, 30,39 Boost, 1,2,5,6,10,16 Brake squeal, 41,42 Carburetor, 89,93 Combustion, 2,4,59-62,64-84,87,89-95, 98,100,108,115, 118,119 crude approximation method, 46,50 CVT, 23-31,33,34,36,38,39 Cylinder block, 105,106,107,109,116 disc brake model, 41,43,44,47 EGR, 1,17,60,61,82-87 Emission, 1-4,6,8,9,12,13,14,17,18,59-61,64-68,71-75,77-87, 89,90,94, 97,98,100, 102,105 FMEA, 106,110,119 Fuel injection, 59,61,67,71,72,93,96,97,98,102

Gasoline engine, 2,20,98,102 Gray cast iron, 105,106,108 High Turbulence, 59,60,62,65,66,69,71,75,82,83,84,87 Identification, 121,122,128,129 neural network, 121,122,125,126,131 Nox, 13,14,60,61,64,66,67,68,71,72,75,82,84,86,87 particle swarm optimization, 122,127 PD controller, 24,29-34,36,39 Pilot injection, 59,61,69-82,84,86,87 quarter car suspension, 122,123 RCS, 119 Robust, 24,43,45,47,55,57 Stepped-piston, 89-93,102

137 Index

test rig, 31,32,110,111,112,123,134 Turbochargers, 2,4-7,12,16,19,20 Two Stroke, 89,90,91,98 V-belt, 27 VGT ,1,2,5-9,13,16,17,18,20

Vibration, 41-45,47,50,52,55,57,98 Ziegler-Nichols formula, 30,39 Zircon sand, 110,111,112,114

research on vehicle technologies_u08

Documents

malaysia oleh

malaysia diterbitkan

vehicle technologies

skudai johor darul tazim

taman universiti

editor srithar rajoo

powertrain engineering

industrial engineering