iv

PISTON RING ASSEMBLY OF A NEW SYMMETRICAL

MULTI-STAGE WOBBLE-PLATE COMPRESSOR

ANDRIL ARAFAT SUHASRIL

A thesis submitted in fulfillment of the

requirements for the award of degree of

Master of Engineering

Faculty of Mechanical Engineering

Universiti Teknologi Malaysia

MAY 2008

vi

To my beloved wife, son and all my big family

vii

ACKNOWLEDGEMENTS

All praises to The Almighty Allah SWT for making everything possible. In

preparing this thesis, I was in contact with many people, researchers, academicians,

and practitioners. They have contributed towards my understanding and thoughts. In

particular, I wish to express my sincere appreciation to my supervisors, Prof. Dr. Md

Nor Musa and Prof. Dr. Ir. Wan Ali Wan Mat, for encouragement, guidance,

contribution and friendship. Without their continued support and interest, this thesis

would not have been the same as presented here.

I am also very thankful to En. Ainullotfi Abdul Latif and all the Researchers

and Research Officers in the UTM Compressor Research Group for their guidance,

advices and motivation. I am also indebted to Universiti Teknologi Malaysia (UTM),

to CNG/DI Engine and Transmission Program, to Ministry of Science Technology

and Innovation (MOSTI) for support and funding my study as a whole.

I would like to thank Dr. Ir. Henry Nasution, MT, Ardiyansyah Syahrom, ST.

MEng, Zair Asrar Ahmad, MEng, Ir. Saiful Jamaan, MEng, Ir. Oktaviandri, MT for

all encouragement and support. As same appreciation is extended to all my

colleagues and others who have provided assistance at various occasions.

Unfortunately, it is not possible to list all of them in this limited space.

I am very grateful to my family, especially thank my wife Era Triana, ST,

MSc and my son Muhammad Alfatih Arafat, for love, prayer, patience and

assistance. As same appreciation is extended to my parents and in-laws, brothers,

sister and to all my big family members for all of their love and support.

viii

ABSTRACT

A new compressor design for a natural gas Home Refueling Appliance

(HRA) as an equipment for a slow fill of Natural Gas Vehicle (NGV) usage has been

developed. It is known as the Symmetrical Multistage Wobble-plate Compressor.

This compressor was designed for a high compression gas that requires oil-free

piston ring assembly in the cylinder to minimise contamination in order to maintain

the purity of the natural gas. The goals of this research were to develop, design,

analyse, and test an oil-free Piston Ring Assembly (PRA) for this new compressor.

Through this research, the material selection and process of designing the piston ring

assembly and considerations taken for this new compressor were explained. To

function as an oil-free piston ring material, Polytetraflouroethelyne (PTFE) and

Polyetheretherketone (PEEK) as polymer material was selected. The performances of

these materials were evaluated using a laboratory scale reciprocating sliding test.

Finite Element Analysis (FEA) was used as an effective tooling to analyse the

component design of the piston ring assembly numerically. Symmetrical multi-stage

wobble-plate compressor prototype with the proposed piston ring assembly was

developed and tested successfully on a test rig which was built for this purpose. The

results met the design requirement with 3 bar of suction pressure, 260 bar of

discharge pressure (more than the specified discharge pressure of 250 bar), 1 m3/hr

of flow-rate, and 1000 rpm of rotational speed. The tests and results presented in this

study were preliminary tests by using air as the working fluid to replace the natural

gas in compression process. The results of this test in this study showed that the

proposed piston ring assembly is able to produce and withstand the extremely high

pressure of about 250bar. The real tests processes on natural gas were conducted

utilising all the experiences and lesson learnt from that of air.

ix

ABSTRAK

Satu rekabentuk pemampat baru untuk Aplikasi Pengisian Semula Kediaman

(HRA) yang merupakan sebahagian daripada peralatan pengisian semula kenderaan

gas asli (NGV) telah dibangunkan. Rekabentuk baru ini dikenali sebagai pemampat

plat wobal simetri berbilang peringkat. Pemampat ini direkabentuk untuk

pemampatan gas asli bertekanan tinggi yang memerlukan penggunaan gegelang

piston bebas minyak di dalam cylinder pemampatan untuk mengurangkan

pencemaran dan mengekalkan ketulenan gas asli tersebut. Objektif kajian ini adalah

pembangunan, rekabentuk, analisa serta ujikaji terhadap gegelang piston untuk

kegunaan rekabentuk pepampat baru tersebut. Di dalam kajian ini juga, proses

pemilihan bahan, merekabentuk gegelang piston dan pertimbangan yang diambil

untuk pemampat baru ini diterangkan. Dua bahan polimer yang berfungsi untuk

gegelang piston bebas minyak telah dipilih iaitu Polytetraflouroethelyne (PTFE) dan

Polyetheretherketone (PEEK). Prestasi bahan-bahan ini ditentukan dengan

menggunakan ujian gelinciran ulang-alik berskala makmal. Analisis Kaedah Unsur

Terhingga (FEM) telah digunakan untuk menganalisa rekabentuk gegelang piston

secara analitikal. Pemampat yang dilengkapi dengan rekabentuk gegelang piston

bebas minyak yang dicadangkan ini telah dibangunkan dan diuji dengan pelantar

ujian yang telah dibina. Hasil ujikaji memenuhi keperluan rekabentuk iaitu tekanan

masukan sebanyak 3 bar, tekanan keluaran sebanyak 260 bar (melebihi keperluan

rekabentuk iaitu 250 bar), kadar alir sebanyak 1 m3/hr dan halaju operasi sebanyak

1000rpm. Ujikaji dan keputusan yang diberikan di dalam kajian ini adalah

merupakan kajian awalan di mana udara digunakan bagi menggantikan gas asli di

dalam proses pemampatan. Keputusan ujikaji di dalam kajian ini menunjukkan

bahawa gegelang piston yang dibangunkan ini mampu untuk menahan tekanan tinggi

sehingga 250bar. Proses pengujian sebenar menggunakan gas asli akan dapat

dipandu daripada pengalaman dan contoh pengujian menggunakan udara ini.

x

CONTENTS

CHAPTER TITLE PAGE

FRONT PAGE i

DECLARATION ii

TITLE COVER iv

DECLARATION v

DEDICATION vi

ACKNOWLEDGEMENTS vii

ABSTRACT viii

ABSTRAK ix

CONTENTS x

LIST OF TABLES xiii

LIST OF FIGURES xiv

LIST OF SYMBOLS xvii

LIST OF APPENDICES xix

LIST OF ABBREVIATIONS xx

1 INTRODUCTION 1

1.1 Background 1

1.2 Statement of Problem 3

1.3 Objective of Research 4

1.4 Scope of Research 5

1.5 Contribution of Research 5

1.6 Thesis Outline 5

xi

2 LITERATURE REVIEW 8

2.1 Introduction 8

2.2 Oil Lubricated and Non Lubricated Cylinder 9

2.3 History and Development of Oil-Free Reciprocating

Compressor 13

2.4 Material development for Oil –Free Piston Ring 17

2.5 Piston Rings Design 24

2.6 Wear of Piston Ring 26

2.7 Computer Modeling and Stimulation 28

2.8 Design Study of an Existing Computer Model 29

3 THEORY OF PISTON RING ASSEMBLY 33

3.1 Material 33

3.1.1 Techniques of Material Selection 33

3.1.2 Piston Ring Design and Material 35

3.1.3 Piston Design and Material 36

3.1.4 Cylinder Liner Design and Material 38

3.2 Design and Analysis 38

3.2.1 Loads and Forces Acting On Piston Ring 39

3.2.2 Elastic of Piston Ring 41

3.2.3 Piston Ring Forces 46

3.2.3.1 Inertial Force 47

3.2.3.2 Pressure Force 47

3.2.3.3 Friction Force 48

3.3 Ring Flutter 48

3.4 Ring Collapse 49

3.5 Gas Leakage on Piston Rings 52

3.6 Wear of Piston Ring – Cylinder Liner 55

3.6.1 Surface Texture of Piston Ring 57

3.6.2 Wear Mechanism 60

3.7 Symmetrical Multistage Wobble-plate Compressor 63

3.8 Installation of Piston Ring 66

xii

4 EXPERIMENTAL METHODOLOGY AND

PROCEDURES 69

4.1 Introduction 69

4.2 Tribotest 69

4.2.1 Material and Specimen Preparation 71

4.2.2 Experimental Procedures 72

4.3 Prototype Experimental Test 73

4.3.1 Prototype and Rig Development 74

4.3.2 Experimental Procedures 76

5 RESULTS AND DISCUSSIONS 77

5.1 Introduction 77

5.2 Tribotest 77

5.2.1 Summary and Discussions 84

5.3 Design and Modeling Using Finite Element

Analysis (FEA) 85

5.3.1 Summary and Discussions on Piston 92

5.3.2 Summary and Discussions on Cylinder Liner 93

5.3.3 Summary and Discussions on Piston Rings 93

5.4 Prototype Experimental Result 94

5.4.1 Summary and Discussions 96

6 CONCLUSIONS AND RECOMMENDATIONS 98

6.1 Conclusions 98

6.2 Recommendations for Future Research 99

REFERENCES 100

APPENDICES 105

xiii

LIST OF TABLES

TABLE NO. TITLE PAGE

2.1 Lubricated Versus Oil-Free/Non-Lubricated (Hanlon, 2001) 12

2.2 Common Examples of Oil-Free Gas Compressor Application

(Wilson, 2000) 16

2.3 Some Field Evaluation of Piston Rings Performance for Some

Types of Oil-Free Compressor on Various Gases

(Wilson, 2000) 18

3.1 Home Refueling Appliance (HRA) Compressor Design

Requirement 63

4.1 Specification of Symmetrical Wobble Plate Compressor 75

5.1 Weight Loss of PTFE Material vs XW41 Hard Chrome

Coated (gram) 82

5.2 Weight Loss of PEEK Material vs XW41 Hard Chrome

Coated (gram) 82

5.3 Results for Stress Analysis on the Piston 86

5.4 Results for Stress Analysis on the Cylinder Liner 88

5.5 Results for Stress Analysis on the Piston Ring Assembly

(PTFE) 90

5.6 Results for Stress Analysis on the Piston Ring Assembly

(PEEK) 90

xiv

LIST OF FIGURES

FIGURE NO. TITLE PAGE

1.1 Flowchart of Research Phases 7

2.1 Typical Piston Rings Ring Assembly of Lubricated Cylinder

Construction (Bloch and Hoefner, 1996) 11

2.2 Typical of Oil-Free Piston and Ride Rings

(Bloch and Hoefner, 1996) 13

2.3 Early Version of Oil-Free Compressor Piston

(Bloch and Hoefner, 1996) 14

2.4 Transfer Film Mechanisms (Dwivedi, 1990) 23

2.5 Scanning Electron Microscopic (SEM) of PTEE Transfer

Film (Dwivedi, 1990) 24

2.6 Typical of Scotch Yoke Compressor 29

2.7 Plunger Piston of Scotch-Yoke Compressor 30

3.1 Flowchart of Material Selection 33

3.2 Scotch-Yoke Compressor and PTFE Added Filter as Piston

Ring 35

3.3 Typical Piston Design (Bloch and Hoefner, 1996) 37

3.4 Cross-Section View of Two Pieces Piston Ring Assembly

in its Sealing Position (SAE, 1969) 40

3.5 Gas Pressure on Two Pieces Piston Ring (SAE, 1969) 41

3.6 Free Body Diagram of a Ring Considering Elastic Ring

Tension Only 42

3.7 Plane Elastic Ring with Uniform Pressure 44

3.8 Piston Ring with Uniform Elastic Ring Pressure 45

xv

3.9 Force Balance Condition for Ring Collapse 50

FIGURE NO. TITLE PAGE

3.10 Three Possible Paths of Gas Leakage Through the Rings 53

3.11 Various Types of Cut Joints for Piston Rings 55

3.12 Rubbing of Two Contact Surface Under Microscopic View

(Chui, 2001) 56

3.13 Surface Profile (Chui, 2001) 57

3.14 Two Different Surface Profiles (Chui, 2001) 58

3.15 Surface Representation Using Abbott Firestone Curve (AFC)

(Chui, 2001) 59

3.16 Small Section of Piston Ring Sliding on Cylinder Liner 61

3.17 Ring Section Contact with Cylinder Liner 61

3.18 Piston Rings in Slices Partition 62

3.19 Cross Section of the Symmetrical Multi-Stage Wobble-Plate

Compressor 65

3.10 Cross-Section of Piston Assembly (First Stage Piston) 66

3.21 Installation processes of the un-cut piston ring

into the piston groove 68

4.1 DUCOM Reciprocatory Friction and Wear Test

Monitoring Machine 71

4.2 Surface Roughness Test For Cylinder Liner

(Material XW41 Hard Chrome Coated) 72

4.3 First prototype of symmetrical wobble-plate compressor

Home Refuelling Appliance (HRA) 74

4.4 Experimental Rig of HRA 76

5.1 Transfer Film Phenomenons 78

5.2 Coefficient of Friction of PTFE Vs XW41 Hard Chrome

Coated 79

5.3 Coefficient of Friction of PEEK Vs XW41 Hard Chrome

Coated 79

5.4 Wear of PTFE Vs XW41 Hard Chrome Coated 80

5.5 Wear of PEEK Vs XW41 Hard Chrome Coated 80

xvi

5.6 Temperature of PEEK vs XW41 Hard Chrome Coated 81

5.7 Humidity of PEEK vs XW41 Hard Chrome Coated 81

5.8 Weight loss of pin (PTFE) 82

5.9 Weight loss of plate (PTFE) 83

5.10 Weight loss of pin (PEEK) 83

5.11 Weight loss of plate (PEEK) 84

5.12 Loads and Boundary conditions of the Piston (First Stage) 86

5.13 Von Mises Stress of the Piston (First Stage) 86

5.14 Deformation of the Piston (First Stage) 87

5.15 Factor of Safety of the Piston (First Stage) 87

5.16 Loads and Boundary Conditions of the Cylinder Liner

(First Stage) 88

5.17 Von Mises Stress of the Piston (First Stage) 88

5.18 Deformation of the Cylinder Liner (First Stage) 89

5.19 Factor of Safety of the Cylinder Liner (First Stage) 89

5.20 Loads and Boundary Conditions of the Piston Ring (PTFE) 90

5.21 Von Mises Stress of the Piston Ring (PTFE) 91

5.22 Deformation of the Piston Ring (PTFE) 91

5.23 Factor of Safety of the Piston Ring (PTFE) 92

5.24 Graph of Pressure vs Time, First Test 94

5.25 Graph of Pressure vs Time, Second Test 95

5.26 Graph of Pressure vs Time, Third Test 95

5.27 Graph of Pressure vs Time, Fourth Test 96

xvii

LIST OF SYMBOLS

A - Piston ring surface area

Ar - Area of cross-section of the piston ring

a - The axial acceleration of the piston ring

am - Moment arm on the piston ring

D - Bore diameter

d - Bore diameter minus two times of the ring width

Em - Modulus elasticity

Fl - Inertial force

Ff - Friction force

I - Area moment of inertia of connecting rod

Iz - Area moment of inertia of piston ring

G - Modulus of elasticity in shear

Mz - Bending moment on the piston ring

m - Mass of the ring

N - Normal force on piston ring

n - Free gap ring

P - Pressure

Pe - Elastic ring pressure at outer radius

Pm - Elastic ring pressure at mean radius

p - Force, which put the ring into loaded shape

q - Loaded ring gap

rm - Piston ring mean radius

rm - Piston ring outer radius

xviii

t - Piston ring thickness in axial direction

Ub - Piston ring strain energy due to bending

Un - Piston ring strain energy due to normal force

Us - Piston ring strain energy due to shear force

V - Shear force on piston ring

x, y - Rectangular coordinates

γ - Piston ring end gap angle

υ - Poisson ratio

δb - Deflection of piston ring due to bending force

δn - Deflection of piston ring due to normal force

δs - Deflection of piston ring due to shear force

δt - Total deflection of the piston ring

µ - Kinematics coefficient of friction

υax - Piston ring in axial velocity

xix

LIST OF APPENDICES

APPENDIX TITLE PAGE

Appendix A Piston Groove Dimension of Piston Ring

(According to ISO 7425/1) 105

Appendix B Piston Groove Dimension of Rider/ Guide Ring

(According to ISO 7425/1) 107

Appendix C Material properties of Piston Ring Assembly (PRA) 109

Appendix D Patent Filing No: 2005 5056

Wobble plate Compressor 113

xx

LIST OF ABBREVIATIONS

CNG - Compressed Natural Gas

NGV - Natural Gas Vehicle

HRA - Home Refueling Appliance

PRA - Piston Ring Assembly

TDC - Top Dead Centre

VRA - Vehicle Refueling Appliance

CHAPTER I

INTRODUCTION

1.1 Background

In response to high petroleum price and environmental concerns, natural gas

becomes an alternative fuel in the market today, as well as solving out of

environmental issues on the higher emission of gasoline uses. Compressed Natural

Gas (CNG) offers the fuel cost-savings to the vehicle owners, due to better efficiency

of energy resource utilisation, provides cleaner burning fuel. Therefore, research in

utilising the gas as another alternative fuel should be given a special priority.

The total energy use by Natural Gas Vehicles (NGV) includes not only direct

vehicle consumption but also the whole processes of extraction, processing,

transportation, distribution and compression of the gaseous fuel. If more natural gas

refueling stations are built, will create more convenience for public to use this new

fuel. In line with this, there are many natural gas vehicle-refueling stations being

built by PETRONAS. Sdn. Bhd (Malaysian national petroleum company). From year

2000 to 2007, there are 39 NGV refueling stations being built nationwide

(http//:www.petronas.com.my). However, this development is not fast enough since

2

the government is targeting for 94 refueling stations by the year 2009 to serve a total

of 57,000 NGV demands.

Since NGV relates to the either distribution or public use of refueling station,

therefore for storage purposes the gas must be compressed to a higher pressure

generally ranging between 200 bars to 250 bars in the refueling storage tanks. The

high compression system is necessary for the natural gas refueling appliances, and

therefore a reliable compressor would be among the main and important equipment

in the refueling facilities.

A new high-pressure multistage wobble-plate gas compressor design for

Home Refueling Appliance (HRA) as equipment of the slow fill of the NGV usage

has been developed. The design is based on a type of the reciprocating wobble-plate

compressor. Basically, this compressor has two compression sides with four stages in

each side of the compression and the overall pressure to be achieved is 250 bars.

Currently, the available wobble plate compressors are only single-sided and single

stage compression, having discharge pressure around 20 to 30 bar and much more

popular in automotive air-conditioning system application. Whereas for this new

design, two sets of wobble plates-piston assemblies were being installed on a rotating

shaft and in a mirror-image arrangement. Further, the multi-stage compression

system has been configured in order to enable the compressor to compress gas to a

very high pressure.

In this new symmetrical multistage wobble plate compressor, an oil-free

lubrication system is one of the specific requirements so that no contamination will

occurs. Since the performance and emission level of natural gas-fueled vehicles are

sensitive to the oil carried-over in the compressor, thus the prevention of

contamination during compression process is very important.

Oil-free or non-lubricated piston rings as a part of assembly compression

system were selected for the compressor in order to achieve a minimum

contamination. However, until now, no literature has been found discussing specific

issue on how to design such an oil-free lubrication system for the wobble plate

3

compressors. Therefore, in this research the focus is on the process of designing the

oil-free piston rings assembly.

1.2 Statements of Problem

Currently, NGV refueling stations nationwide are installed with imported

models of reciprocating gas compressor. This compressor usually uses oil as

lubricant inside the crankcase and cylinder wall, where all friction parts are

lubricated with oil. In another gas compressor model, oil mist is used to lubricate the

piston rings controlled with a timer. In this type of compressor, the final discharge is

freed from traces of oil by using separators and filters.

Nevertheless, the oil is normally not fully removed. In addition, the effect of

“trapped-oil” could contaminate the gas inside the tank, it could affect compression

process and dropped the combustion performance of the engine. To overcome this

problem, the solution is to use such compressors that operate without any lubrication

oil (oil-free) especially on the inside wall of the cylinders.

Conceptually, a criteria of success for the compression process in oil-free

compressing gas depends on the piston ring assembly design. Several factors such

geometry, material selection, friction, wear and tribological influences are important

parameters to design and to make the analysis ensuring the “good sealing”

compression process for this new compressor. The good sealing means low leakage,

low friction, low power consumption, low wear, low temperature rise, long life

operation and high efficiency of compressor.

Piston rings for current reciprocating compressor have to meet all the

requirements of a dynamic seal for linear motion that operates under demanding

tribological conditions. During sliding process between piston ring and cylinder liner

cause the friction and wear. Piston rings assembly wear would occur on the contact

surface between ring, piston and cylinder wall after a certain amount of time of

4

operation. Due to the operation within, the contact surface usually experiences much

higher pressure than other parts and gets much force, which would cause deformation

on the geometry and degradation of quality of the surface material. Then eventually

the distorted surface affects the functionality of the piston rings, and results

significant energy losses. Because of that wear, the piston rings also lose their sealing

function. To overcome these problems, the piston ring material should be selected

particularly one that has small thermal coefficient of expansion, good creep

resistance, good resistance to chemical attack to prevent any gas leakage.

Examination of performance on the new compressor was conducted to have a

very high gas pressure exerted in the final stages rise up 250 bar. During the

operation of compressor with high pressure difference across piston rings contributed

significantly to the ring extraction between the piston and cylinder liner clearances.

On other hand, lowering clearances reduced the ring extraction, but increased the

possibility of piston contact with cylinder liner, while the piston rider rings also

being worn out. Further, high pressure also generated higher surface contact

temperatures. This temperature it higher than the measured gas discharge result in the

piston ring creep and extrusion.

At the other challenge, small final stages piston diameter (10 mm) and piston

ring geometry at very high pressure also needed a consideration to assemble more

pieces of ring in order to prevent gas leakage. Eventually, by placing more rings the

friction force will increase, temperature, power consumption and wear rate will also

increase. Therefore, to reduce all these affects a careful selection of material, design

and analysis of piston ring assembly are very critical in this new developed of high

pressure symmetrical wobble-plate compressor.

1.3 Objective of Research

To develop an oil-free piston ring assembly for a new multistage symmetrical

wobble plate compressor.

5

1.4 Scope of Research

The development work of the new symmetrical wobble-plate compressor was

carried out by a team of researchers and each member has a scope to focus upon. For

the author’s scope, this research was focused on the overall development of Piston

Ring Assembly.

1.5 Contribution of Research

The contributions of this research were developing a new piston shape,

cylinder liner shape for a new symmetrical wobble-plate compressor.

1.6 Thesis Outline

The thesis outline is divided into five stages. The first stage is concept

development for the oil-free piston ring assembly based on literature review. This

includes a through understanding of the problem by going through the literature

review, and reserve engineering work. In the development of the existing oil-free

piston ring assembly, it is found that basically a reciprocating compressor using a

vertical, horizontal (in opposed design) and scotch yoke mechanism are using a

crankshaft with crosshead mechanism to transfer the movement of the piston which

slides in and out of the cylinder. For specific comparison, of an oil-free compressor

the Balance Scotch Yoke mechanism was studied and reverse engineered.

In the second stage of the project, theory of piston ring assembly was carried

out. It was done by taking various references from the existing oil-free compressor,

and comparison from such an established manufacture of the sealing materials. Some

technique selections of material were conducted in this research such as imitative and

comparative procedures. Considering the material characteristic in high temperature,

a selection of the oil-free material of piston ring assembly based on polymer resin

6

such as polytetrapolyethelyne (PTFE) and polyetheretheleneketone (PEEK) were

adopted in this research.

In the fourth stage of the work a laboratory scale tribotest was also conducted

to establish the characteristic of the material selected. A reciprocating wear method

was used to measure the friction and wear rate to predict the life span of the

contacting piston rings-cylinder liner. The results in the experiment give the real

value of the coefficient of friction of sliding parts, wear coefficient, and figures out

the film transfer phenomena between piston ring-cylinder liner as well as to know the

type of wear that happens during sliding. At the same time, the surface roughness

affect and other tribological aspects were also studied. This stage also describes the

modeling and simulation method of the piston ring assembly were using

Computational Aided Design (CAD) software and Finite Element Method (FEM)

approach. The computational static analysis was used to check the sizing geometry

and material performance, to ensure that the part would not fail. Von Misses (stress)

and deformation value for each part were calculated and compared with yield

strength of material to obtain the part safety factor. To do all these, using a

commercial Solid-Work integrated with COSMOS Finite Element software. From

these analyses, the piston ring assembly design parameters and its relationship are

revealed.

The final stage was the development of the prototype and the rig followed up

by discussion of the experimental results. The focus of this experiment was to

monitor the performance of actual designed piston rings. The main objective in this

test was to verify the performance of piston ring at the specified pressure of up to

250 bar. The test was also equipped with Data Acquisition System (DAS).

7

Literature Study

Concept Development

Reverse Engineering

Design of Piston Ring Assembly (PRA)

Figure 2.1 Flowchart of Research Phases

END

Design & Analysis Tribotest Material Selection

Modeling & Simulation

Development Prototype & Rig

Prototype Test

START

Data Analysis

CHAPTER II

LITERATURE REVIEW

2.1 Introduction

In a reciprocating compressor, the system of piston rings assembly is one of

the most important and critical aspect. Approximately 60% of the frictional forces

caused in the reciprocating machines are the result from this tribological system of

piston ring as reported by Todsen and Kruse (1982). In order to achieve efficient

sealing, the piston ring should make a good fit with both the cylinder wall and the top

or bottom of the piston ring groove. Piston rings for current reciprocating compressor

have to meet all the requirements of a dynamic seal for a linear motion that operates

under demanding pressure, thermal and chemical conditions. In general, the

following requirements for a set of piston rings assembly can be identified as:

• Low friction, for supporting a high power efficiency rate

• Low wear of the ring, for ensuring a long operational lifetime

• Low wear of the cylinder liner, for retaining the desired surface texture of the

liner

• Emission suppression, by limiting the flow of crankcase oil to the combustion

chamber

9

• Good sealing capability and low blow-by for supporting the power efficiency

rate

• Good resistance against thermal, chemical attacks and hot erosion

• Reliable operation and cost effectiveness for a significantly long period at

time

2.2 Oil Lubricated and Non Lubricated Cylinder

Since the cylinder assemblies of reciprocating compressors must be designed

relative to their lubrication, the nomenclature used to describe and classify the types

of cylinder construction likewise refers to lubrication. The classifications most

commonly used are depicted from Bloch and Hoefner (1996):

(i). Lubricated Cylinder Construction.

The lubricated cylinder assembly is the conventional cylinder construction,

which has a liquid lubricant introduced directly into the cylinder and piston

rod packing in sufficient amounts to provide a lubrication film between the

mated parts. The gas from the lubricated cylinder is contaminated with the

lubricant, normally a hydro-carbon or a synthetic oil.

(ii). Mini-Lube

A partially lubricated cylinder construction with oil feed to the cylinders

reduced to at least one-third of that for a lubricated cylinder. Teflon as self-

lubricated material is used on the piston and for the pressure packing. The

aims of Mini-Lube construction are to reduce the amount of oil carried within

the exit gas and to reduce contamination of systems.

(iii). Micro-Lube

No lubrication to the cylinder from conventional oil feed, but some oil enters

the cylinder from migration along the piston rod. Teflon as self-lubricated

material is used on the piston and for the pressure packing. The oil is usually

10

removed by scraper rings, which allows oil migration along the piston rod.

The reasons for this construction are the same as for Mini-Lube, except that

the system receives an even smaller amount of oil.

(iv). Non-Lube or Oil-Free Cylinder Construction

No lubrication reaches the cylinder. A longer distance piece between piston

rod and cylinder is used to separate the crosshead guide from the cylinder.

This necessitates a longer piston rod on which a "collar" or oil deflector is

installed. This collar prevents oil migration along the rod and into the

cylinder.

The conventional lubrication of piston ring in an industrial gas compressor or

combustion engine usually used lubricated cylinder construction as explained above.

Oil as lubricant functions to reduce the friction-wear between piston ring and

cylinder liner. Oil also has the functions as a media to assist the transfer of heat from

piston to cylinder wall and to control oil consumption.

Piston ring assembly forms a ring pack, which usually consists of 2–5 rings,

including at least one compression ring. The number of rings in the ring pack

depends on the engine type, but usually comprises 2–4 compression rings and 0–2 oil

control rings. For example, fast speed four-stroke diesel engines have 2 or 3

compression rings and a single oil control ring. The oil control rings used in diesel

engines are two-piece assemblies and the spark ignited engine of oil control rings can

be of three-piece assemblies. In addition to the general compression rings and oil

control rings, there are scraper rings which have the tasks for both sealing and

scraping off the oil from the liner wall, see the Figure 2.1.

Many applications in industrial gas compressor, the oil that sips into the gas

flow system is generally acceptable but, equally, there are a wide range of uses of

compressed gases (in the food industry, brewing and pharmaceutical industry, for

breathing air, chemical and petroleum industry, etc) where the presence of

lubricating oil is completely unacceptable.

11

Figure 2.1 Typical piston rings ring assembly of lubricated cylinder construction (Bloch and Hoefner, 1996)

The oil contamination during compression can create the sludge that surely

will reduce performance of a compressor and in some cases can possibly lead to a

combustion of the system if the oil is passed into the machine. For these and other

reasons, oil-free cylinder construction has become increasingly popular as describes

by Bloch and Hoefner (1996). Further brief discussions on the advantages and

disadvantages of oil lubricated compared to oil-free/ non-lubricated compressor are

given in Table 2.1.

In the non-lubricated or oil-free construction, piston and piston ring assembly

there is no oil film to wet the piston, so the metallic piston must be kept off the

cylinder bore by other means or else serious damage will result. Note that this is the

difference between lubricated and non-lubricated principle.

Consequently, the material for the oil-free piston ring must have certain

characteristics to fulfill the function of piston ring as have been explained before.

The popular materials used in oil-free compressor application are carbon, graphite

and Polytetrafluoroethelyne (PTFE). These materials usually called as the self-

lubricating material, where by the process of steady wear can release a loose carbon/

12

graphite material which acts as a lubricant between the piston and liner. This

phenomenon will be described further in the next section.

Table 2.1 Lubricated versus oil-free/ non-lubricated (Hanlon, 2001):

Lubricated Oil-free/ Non-lubricated

Advantages Increased piston ring life Low to nil contamination of discharged

gas Allowed use of metallic ring Reduced overall lubrication requirements Air cooling or non-cooling system Less discharged gas filtration needed

Higher pressure ratios and discharge

temperature Reduced routine maintenance

Fewer stages necessary in some case

Higher operating speed Reduced capital cost Longer overhaul intervals Disadvantages Oil contamination of discharge Higher maximum discharge temperatures

Oil deposits in pressure vessels reducing

capacity to store gas Reduced piston ring and rod packing life

Oil contamination of on board vehicle

equipment Lower pressure ratios Increased vehicle emissions More stages may be necessary Higher compressor oil consumption Increased capital cost

Increased maintenance on lubrication

system Lower operational speeds

Increased noise levels with air cooled

compressors

In order to ensure the durability, there are in some design where the oil-free

piston rings are to serve a large cross-sectional area, it consists a minimum of two

piston rings and one rider ring. The rider ring acts as the support piston weight and as

a bearing surface to transmit the piston side loading into the cylinder wall. A typical

piston rings assembly for oil-free compressor is equipped completely with piston

rings, rider rings, and rod packing. Figure 2.2 illustrates and shows the location of

these critical components.

13

Figure 2.2 Typical of oil-free piston and rider rings (Bloch and Hoefner, 1996)

In the conventional non-lubricated compressor, the piston is kept off the

cylinder wall by a guide ring which is referred to as a wear, or rider ring. This rider

ring is of a low friction material, such as carbon or Teflon, and of low unit loading

relative to the piston weight. The outside diameter of this piston ring is smaller than

that of the piston in the lubricated compressor model, this creates clearance between

the piston outside diameter and the cylinder bore. This clearance allows for rider

band wear before metal contact occurs with the cylinder bore. The rider ring is either

a solid or a split configuration; its size is determined by piston assembly weight only

and is independent of operating pressures.

2.3 History and Development of Oil-Free Reciprocating Compressor

Bloch and Hoefner (1996) described the history of oil-free compressor in the

industrial compressor. Around mid 1930s, the first high pressure 2000 Psi oil-free air

compressor was made by using carbon piston rings used the water as lubrication. In

following years, many single and multi stage compressors were made by using

carbon as the wearing material for both piston rings and rider rings. This carbon

piston ring construction is shown in Figure 2.3. This was a “non-floating” type

14

piston, which means that the carbon rings transferred the weight and load of the iron

piston onto the cylinder liner. Piston rings with expanders were used to seal the gas.

Another type of construction was a “floating” piston, in which a tail rod was

used with a small auxiliary crosshead. The tail rod supported the piston and

prevented it from touching the cylinder liner. Carbon rider rings were not used. The

carbon material is an extremely brittle and requires extreme care when it is installed

to prevent breakage. The carbon dust generated as a result of wear is somewhat

abrasive and accelerates further the ring wear. Ring slap, caused by the another

resulting excess side-clearance of the rings in the grooves, tends to chip or fracture

the rings.

Figure 2.3 Early version of oil-free compressor piston (Bloch and Hoefner, 1996)

Poole (1978) summarised development industrial oil-free compressors in the

beginning of the 19th century (1950 – beyond 1978). Industrial oil-free compressor

used in this decade has grown more increasingly particularly since 1950 compared to

that of the lubricated compressor. Many applications used compressor as unit of

service that required compression discharge without oil contamination. There are

wide ranges of use of compressed air/ gas for instrument and control purposes, such

as in the food industries, brewing, pharmaceutical industries, and for breathing.

Carbon and graphite are popular materials for oil-free compressor cooled by air and

15

water. Because of the safety and environmental aspects before 1950, the compressor

just operated no higher than 7-8 bar.

From the use of compressor with a single stage, single acting and double

acting with multistage compression, the later compressor manufacture has gradually

accepted more and more of the responsibility for the total installation. Especially in

industrial air compressor, the engineering development of oil-free type compressor

with higher reliability is also reported. With the use of air-drying equipment to

extract the water vapor remaining inside the delivered air from-and-after the cooler,

the manufacturer can provide the package of compressed air installation that delivers

air freely from any significant contamination and they were capable of operating

with minimum maintenance and maximum reliability.

Since the 20th century, construction and development of oil-free compressor

have been challenged by international requirement of the recognized standard of

American Petroleum Institute (API 618) “Reciprocating Compressors for Petroleum,

Chemical, and Gas Industry Services” dated June 2005:

“ The equipment (including auxiliaries) covered by this standard shall be

designed and constructed for minimum service life of 20 years and an

expected uninterrupted operation of at least 3 years…”

Wilson (2000) presented a paper discussed the advanced materials for the oil-

free reciprocating compressor. In this paper, he and Compressor Product Company

(CPI) company’s had developed new materials for oil-free compressor parts

especially for piston, rod packing ring, and guider ring. Investigated on field

compressor applications with various gasses, the successful development of oil-free

compressor answered the API 618 challenge. The common development of oil-free

gas compressor that has been using in industries:

16

Table 2.2 Common examples of oil-free gas compressor

applications (Wilson, 2000):

Gas Application Examples

Ammonia Refrigerant, chemical processing agent

Air PET bottle blowing, air separation, pneumatic instruments

Argon Welding, lamps

Butane Fuel gas, chemical manufacturing

Carbon Dioxide Carbonation of drink, cooling, fire extinguishing

Carbon Monoxide Chemical processing, ore reduction, fuel gases

Ethylene Plastic manufacture, antifreeze

Helium Welding, lamps, cryogenic, balloons

Hydrogen Refining, food manufacture, ammonia synthesis

Isobutene Plastic and chemicals manufacture

Methane Fuel gas, chemicals manufacture

Natural gas Fuel gas

Nitrogen Inert gas purging, ammonia synthesis,

Oxygen Steel and chemicals manufacturing, breathing systems

Propane Fuel gas, refining

Propylene Plastic and chemical manufacturing

From Table 2.2, some of the gases the compressor in a “bone dry” condition.

This bone-dry condition means that the gas demanded a prior drying process

especially if the gas is exposed to a lower temperature during the liquid-gas phase

change. For example, liquid natural gas is stored below -160ºC.

Existing reciprocating compressor, with carbon-filled Polytetrafluoroethelyne

(PTFE) piston rings, guide rings and packing rings fitted to compressor that can

handle those dry gases are summarised in Table 2.2, have been found to be able to

operate between 500 hours to 6000 hours duration before the next maintenance.

Developments of material were introduced by CPI manufacture from special polymer

alloy, (a code CPI 184). Using this material CPI 184 the oil-free compressor can

17

operate with pressure ranging from 0.5 to 36 Mpa and majority of gases are in a

bone-dry condition. Some field evaluations of piston rings performance for some

types of oil-free compressor on various gases are summarised in Table 2.3:

2.4 Material development for oil-free piston ring

Traditionally, piston rings for reciprocating motion were made from the cast

iron. Cast iron is combined with steel as cylinder liner, so that lubrication with oil

becomes necessary to reduce friction and also as part of the cooling system.

Lubrication oil film also has the functions to prevent a leakage between piston ring

and cylinder liner. However, for oil-free reciprocating motion, the oil function is

replaced by the solid lubricant film which can be transferred during the sliding

between piston ring and the smooth surface of the steel cylinder liner. In

reciprocating compressor, the system of piston-ring-cylinder liner or Piston Ring

Assembly (PRA) is important assemblies to be concerned. The high pressure,

temperature, friction, and wear during compression process make PRA becomes very

important for design and should be optimized for achieving a minimum of power and

compression losses.

Maczek and Wolek (1994) investigated the technology of air compression

using oil-free reciprocating compressor in which cylinder and piston rings are

specially designed and modified by manufacturers for a stable operation. Two

compressors of air compression have been redesigned using a cylinder made of

aluminum alloy and cylinder bearing surface that covered by 80 micron-meter

electrolytic oxide layer of aluminum oxide. Piston rings were specially designed and

made from the modified PTFE (15% of graphite and 2.5% molybdenum disulphide).

Air pressure of the both compressor types 1 and 2 specification raised up until 0.2

MPa and 1450 rpm rotational speed. Type 1 is a single stage, twin cylinder confined

lubrication. Whereas for type 2 is single stage, one cylinder and oil-free condition.

18

Table 2.3 Some field evaluations of piston rings performance for some types of oil-free compressor on various gases (Wilson, 2000)

Gas/ Discharge

Pressure (MPa)

Approximate Piston

Ring Lives (Hr)

After Changing with

CPI Piston Ring (Hr) Description Figure

Isobutane

(1.94 MPa)

500-4000

20000 (CPI 184)

Non-lubricated dry isobutene in two

stages to a pressure of 1.94 MPa.

Originally fitted with piston and rider

rings made from epoxy resin bonded

composite and subsequently carbon

filled PTFE. Typical operating lives

500-4000 hours. CPI 184 piston rider

ring indicates life 20000 hours.

Hydrogen

(4 MPa)

2000

16000 (CPI 184)

Non-lubricated dry hydrogen

compressor, which deliver in two

stages to a pressure of 4 MPa. When

installed these were fitted with carbon

and ceramic filled PTFE piston and rod

seals has time for 2000 hours. After

changing with CPI 184 improved to

16000 hours.

19

Gas/ Discharge

Pressure (MPa)

Approximate Piston

Ring Lives (Hr)

After Changing with

CPI Piston Ring (Hr) Description Figure

Natural Gas

(3.5 MPa)

500-1000

16000 (CPI 184)

Compressor for dry natural gas in three

stages to a pressure of 3.5 MPa. Piston

and rod seals made from CPI 184

material provide operating life of

around 16000 hours. Previously used

material (carbon-filled PTFE and epoxy

resin bonded composite) exhibited

variable life spans, down to less than

1000 hours

Ethylene

(2.3 MPa)

2500

4000-8000 (CPI 184)

The compressor handles boil-off

ethylene gas from liquid storage,

raising pressure in three stages to 2.3

MPa. A variety of piston sealing

materials including carbon-filled PTFE

have failed to provide a life of more

than 2500 hours. CPI 184 piston rider

ring indicates life spans 4000-8000

hours.

20

Gas/ Discharge

Pressure (MPa)

Approximate Piston

Ring Lives (Hr)

After Changing with

CPI Piston Ring (Hr) Description Figure

Nitrogen

(2.17 MPa)

2000

16000 (CPI 184)

Non-lubricated dry nitrogen in two

stages to a pressure of 2.17 MPa.

Originally fitted with piston and rider

rings made from epoxy resin bonded

composite and subsequently carbon

filled PTFE. Typical operating life

2000 hours. CPI 184 piston rider ring

indicates life span16000 hours.

21

Bottomley (1994) presented the history of Polytetrafluoroethelyne (PTFE) for

self-lubricated material in reciprocating machine. The history of PTFE began April

6th 1938 to present at Du Pont’s Jackson Laboratory in New Jersey, United States of

America. The main advantage of this material is it can perform a self-lubrication

because the properties of PTFE is based on the process of steady wear, which

releases that loose carbon/ graphite material that acted as a lubricant between piston

and liner. The PTFE piston ring also deposits an adherent counter-surface of PTFE

on to the liner wall during the compression process (adhesion wear). To improve the

performance of PTFE the additives or fillers such glass, carbon, graphite, bronze and

molybdenum disulphide are normally used. By careful formulation and selection of

filler materials, the self-lubrication properties of the filled PTFE materials have been

improved to give the longer life-times, especially in gas compressors. Parallel

detailed analysis using simulation such as Finite Element Method of piston ring

material has also promised more optimization of the selected materials.

To make the material stronger and more resistant against the tendency of

creeping, the fiber or bronze fillers can be added as fillers. Bottomley (1994) also

reported the synopsis of development of the filled polymeric compound. Fillers are

often used in a combination, particularly between carbon and graphite. Carbon for

dry gas usage, at low and cryogenic dew point, is good and useful as addition of

specific fillers. The basic fillers used in the development of oil-free gas compressor

are glass, carbon, graphite, molybdenum disulphide (MoS2) and bronze:

• Glass

Glass which is the most widely used filler is the milled glass fiber. Glass

improves the creep resistance of PTFE at all temperatures. It is chemically stable and

improves the wear rate and friction characteristics of PTFE.

• Carbon

Carbon has excellent resistance to chemical attack, except in oxidizing

environments such as concentrated acid, where glass performs better. Carbon adds to

the creep resistance, increases the hardness and raises the thermal conductivity of

PTFE. In general service carbon filled compounds have more excellent wear

22

properties, particularly when combined with graphite. The combination of the above

properties makes the carbon-graphite filled PTFE becoming a standard choice for oil-

free operation in industrial compressor. During the mid to late of the eighties, carbon

in fiber form began to emerge as a successful filler, particularly in dry gases. Carbon

fiber changes the physical properties in a similar way to glass fiber. Generally,

material that has less carbon fiber than glass fiber is needed to achieve the same

effect. Carbon fiber is chemically inert so that it can be used to replace the glass

filled compounds that fail to resist it. Additional advantages also accrue: higher

thermal conductivity, lower thermal expansion coefficient, and lightness. Carbon

fiber filled materials have less wear during contact with most metals and are less

abrasive on mating surface.

• Graphite

Graphite is a crystalline modification of high purity carbon. It has excellent

wear properties, particularly against soft metals, and display good load carrying

capability in higher speed contact applications. Graphite is also chemically inert, and

can be used in combination with other fillers. Filled PTFE containing carbon and

graphite has one of the lowest coefficient of friction of the filled PTFE compounds.

• Molybdenum Disulphide (MOS2)

Molybdenum Disulphide is used in low percentages and normally only with

other fillers. MOS2 adds to the hardness and stiffness of the PTFE and also reduces

friction.

• Bronze

The additional of high percentage of bronze powder to PTFE has the result of

higher thermal conductivity and better creep resistance than most other filled

PTFE’s. Single stage air compression is a successful example of the bronze filled

PTFE but it should not be used for sour gas applications as the pressure of Hydrogen

Sulphide (H2S) in the sour gas attacks the bronze.

Radcliffe (2005) explained the transfer film of the self-lubricating material.

During the sliding process, the PTFE material is sheared away from the piston rings,

and some of it are deposited into the cylinder liner to form a transfer film, as it is

23

shown in Figure 2.4. Further subsequent sliding takes place between PTFE and

PTFE, for which the coefficient of friction is extremely low. This phenomenon

ensures a considerable longer life for the piston rings as well as for the liner

(Dwivedi, 1990).

Figure 2.4 Transfer film mechanisms (Dwivedi, 1990)

In general, simultaneous process of piston ring wear and formation of transfer

film is a combination of tribochemical and mechanical reaction phenomena.

Depending on the type of fillers, surface roughness of sliding material and gas

conditions, a range of reaction can take place in the transfer film so that the wear

behavior of a filled PTFE material varies with the gas conditions and the fillers used.

To show how the wear affects on the piston ring or the cylinder liner, an image

captured by a Scanning Electron Microscopic (SEM) is shown in Figure 2.5.

24

Figure 2.5 Scanning Electron Microscopic (SEM)

of PTFE transfer film (Dwivedi, 1990)

2.5 Piston Rings Design

The studies on the kinematics and dynamics motion of the piston and piston

ring as function of the liner and piston geometry, surface quality, thermal and

thermodynamic boundaries have been done by Wrede (1978) and Haubner (2001).

Primary and secondary motions of the piston are used as boundary to simulate the

dynamic behaviour of the piston ring. Piston secondary movement is described by

the displacement of the piston normal to the liner and crank shaft axis and the tilt

angle around the pin axis. This secondary movement is influenced by two main

factors, geometrical and operational.

Under geometric factor the secondary movement are determined by the piston

axis deviation, eccentric centre of gravity of piston (near piston pivot point) and

distortion of liner axis to crankshaft axis, whereas gasses and mass forces are the

operational factor. Newer requirement for piston design that need to be addressed

includes light weight design, low friction loses, high wear resistance and lifetime

demands. Piston group is placed separate from the liner by the lubricated oil film.

25

Piston skirt will carry the inclination of the connected rod force. Skirt load

capabilities are determined by grinding of the piston, liner distortion, piston structure

stiffness, surface roughness between the piston and the liner including the liner

honing, crank train kinematics, masses or mass distribution, local liner, oil film

temperatures, oil quality and oil viscosity. Haubner has also listed out the technique

to reduce piston ring friction as follows:

• Increased of cylinder liner temperature

• Reduced of tangential forces in combination with reduced bore distortion

• Optimized ring geometry (friction behaviour is ruled by run in)

• Use of new materials with high resistance, durability and efficiency

• Reduced ring preload

• Reduced ring height

• Special ring design

• Reduced number of compression rings

• Reduced liner distortion

• Increased of liner temperature

• Low oil viscosity

Many studies and investigation have been reported in the literature in the

optimasitation of design at piston rings for engine and compressor. Ouwerkerk and

Theeuwes (1981) have been developed a test rig to determine the leakage and friction

in piston ring-liner of engine and gas compressor. A test rig was used in which the

friction was measured during the stroke and the gas leakage over a whole cycle.

Different shapes with o-ring backup methods and cut joints were tested in this

research. They reported that there were three possible leakage paths; gap between

piston and liner, rear side of the piston ring and joint cut of the piston ring.

Yong (1986) studied a method for predicting the sealing characteristics of

piston rings and evaluated the sealing effects of lubricating by oil and solid. He

developed the mathematical models of working cycle in a cylinder and the gas

leakage through piston rings. In oil-free compressor, there are three possible paths of

gas leakages through the rings; ring and surface of cylinder wall, rings and bottom of

26

slot of piston, and gaps of rings. The mathematical simulation of working cycle of

compressor is used to calculate pressures in cylinder, which has some proper

simplifications:

• Thermal parameters in suction and discharge plenums are constants

• No heat transfer in the cycle

• Gas leakage through piston rings is negligible

• No gas leakage through valves and the flow coefficient of valves are constant

• The working medium is ideal gas

2.6 Wear of Piston Ring

It is commonly assumed that the wear of piston rings proceeded according to

a mild mechanism of mild two-body abrasive wear against the cylinder liner, as

being expressed by the formulae presented by Gupta (2001), Kauzlarich and

Williams (2001) in the reality process of the wear is significantly more complicated.

The wear of piston rings and cylinder liners can be accelerated by three-body

abrasive wear due to the minor abrasive particles in the lubricating oil. The

contaminant particles caused by the three-body abrasive wear can originate from the

oil sump or from the combustion chamber.

For low wear rates, the wear volume of piston rings can be determined by

comparing the surface roughness between before and after the tests on the surface

roughness profiles or cross section profiles, Shuster et.al (1999). Alternatively, the

wear can be estimated from an analysis of the changes at relevant surface roughness

parameters that represented certain proportions of the piston ring face-surface area,

Sherrington and Mercer (2000). For high wear rates, the wear volume can be

determined from macro geometrical changes or mass loss.

Wilson (1990) investigated the materials for oil-free gas compressor,

especially for piston ring and rod packing material. In this paper, Wilson also

27

describes the wear process of self-lubricated materials. The process by the self-

lubricating sealing components (that provided their own lubrication and wear

resistance) can be described as a transfer mechanism. This phenomenon involves a

complex mechanical deposition of two frictional materials. Results in a thin transfer

film of some identical materials become intimately attached into the counter-surface

(cylinder liner). Once this transfer film has been established, the rate of wear of the

component can be relatively stable, ideally it is reduced to almost negligible rates.

Wilson also observed the influence of wear parameters:

• Piston mean sliding speed

• Specific loads (on rider rings)

• Gas pressures and temperatures

• Liner and rod surface finishes

• Cylinder cooling efficiency

• Gas cleanliness

Priest and Taylor (2000) presented the modeling and simulation of piston

ring-cylinder liner wear phenomenon. Wear of both piston rings and the cylinder

liner is perhaps the most difficult phenomenon to be fully implemented in a

calculation model. Wear parameters most certainly require empirical data. The time

required for conducting the simulation increases when the wear models are included

in the simulation software. In terms of simulation, the wear comprises the less

understood phenomena rather than for friction or lubrication. Even though wear

might be considered a minor factor in calculation models, it should be remembered

that the wear of a piston ring alters the ring profile. Therefore, wear is a phenomenon

to be included in a realistic model. Priest and Taylor have investigated piston ring

wear modeling. They pointed out that piston ring designs with emphasis on wear

resistance may be non-optimal if considering the lubrication and frictional properties.

Wear of the cylinder liner at a great extend is caused by the action of the

piston rings (Affenzeller and Gläser, 1996). Practical observations and theoretical

analyses correlate well in terms of the strongest wear of the cylinder liners by taking

place near to the top reversal point of the top piston ring, where the thermal,

28

chemical, erosive and abrasive conditions are the severest. High wear of the cylinder

liner is furthermore associated with the top reversal point of the second piston ring,

and (to a less extent) with the bottom reversal points of the piston rings. Carbon

deposits above the ring pack on the piston may significantly increase the cylinder

liner wear in sliding processes.

2.7 Computer Modeling and Simulation

Dunaevsky (1999) investigated the fiction temperature generated by a piston

ring in a reciprocating oil-less air brake compressor. Many parameters that influence

prediction on the friction temperature are reciprocating motion of the piston, gas

load, piston ring geometry, thermophysical properties of the ring and the bore

material. The three dimensional diffusion-equations conducted to solve the

rectangular source of heat are involved in a reciprocating motion. The solution is

presented in an integral form, and the results are obtained using numerical

integration.

Shivakant and Krishna (2001) have presented a work on simulation of a

piston ring in a multi-body single cylinder internal combustion engine by using FEA.

Several studies, like hydrodynamic lubrication, blow-by, contact between ring and

liner and the piston secondary motions are important for performance evaluation and

design of piston rings. For their simulation model, they studied crankshaft,

connecting rod, piston, piston ring and liner. Explicit method that is well-suited for

condition at high-speed dynamic events, at short duration and at nonlinear contact is

used. By knowing the initial clearance and piston ring deformation, then the blow-by

can be predicted. Blow-by occurs if the ring liner-gap exceeds the oil film thickness.

Axial movement of the piston ring in the groove also contributes to the blow-by.

Fatigue failure needs to be considered since the piston ring is subjected to the cyclic

loading. The clearance between piston and liner is responsible for two secondary

motions, tilt and translational motion of the piston assembly that is perpendicular to

the piston pin and engine axes. The similar method is used by Dunaevsky (2001).

29

2.8 Design Study of an Existing Compressor Model

A typical scotch-yoke compressor for NGV refueling appliance which is

produced by Sankyo Corporation is shown in Figure 2.6. Reverse engineering was

done on this 3m3/hr scotch-yoke compressor. The Scotch Yoke compressor concept

was introduced for small oil-free multi-stage reciprocating high pressure compressor

design developed for Natural Gas Vehicle (NGV) application and other type of

gases. The main features of the yoke compressor are compact design, hermetically

sealed, quiet and low vibration, oil-free compression, operating temperature ranging

from -400C to 400C, easy serviceable and long service intervals (Baumann, 1994).

Figure 2.6 Typical of scotch-yoke compressor

Piston diameter varies from the largest to the smallest from the first stage to

the last stage. The first, second and third stage are equipped with piston rings of

special PTFE-compounds sliding against hard-anodized aluminum of cylinder liner.

Scotch yoke compressor design uses pre-stage compression piston with double acting

piston at first stage. This concept will increase the flow-rate and efficiency of the

compressor and has been used for all stages in automotive air-conditioning scotch

30

yoke compressor ((Riegger, 1990); (Baumann, 1994); (Nishikawa and Nishikawa,

1998 and 2000); and (Bauman and Conzett, 2002)). Pre-stage compression piston is

only suitable for large piston diameter size and unsuitable for small piston diameter

size for second stage and above in multistage scotch yoke compressor.

Some methods have been used to increase the durability in the scotch-yoke

compressor as presented by Nishikawa et al (1998). Grease with good heat resistance

and little scattering property was selected for the sliding parts in the yoke

mechanism. Anti-wear materials which were suitable for the sliding movement in a

non-lubricated condition were used for the piston ring at the lower pressure stages.

For the high pressure stages where the piston diameter size was small and not

suitable for the piston ring usage, plunger piston concept was used. Labyrinth

grooves on the plunger piston surface keeps leakage to minimum. Rolling support

was used at the piston to guarantee the piston free from side-force. The plunger

piston floats freely in the cylinder with a small well defined gap, as can be seen in

Figure 2.7.

Figure 2.7 Plunger piston of scotch-yoke compressor

31

Baumann and Conzett (2002) reported that the clearance gap of 4 to 6 µm in

diameter is needed for the sealing between the plunger piston and the cylinder liner.

The thermal coefficient of expansion for the plunger piston and the cylinder liner

material also need to be small and similar due to this small gap. The materials for

plunger piston and cylinder liner used for gases like air, nitrogen, natural gas and

carbon dioxide were hard-metal and ceramic respectively. Gases like helium, argon,

dry nitrogen and hydrogen require special material combinations and coatings for the

plunger piston-cylinder liner combinations. They have also shown that the leakage

flow through the clearance gap which constitutes a few percent of the gas flow rate is

laminar and depending on the piston geometry, the clearance dimension, the piston

velocity and the type of gas.

CHAPTER III

THEORY OF PISTON RING ASSEMBLY 3.1 Material

This chapter discusses the overall theory of the materials for piston ring

assembly, which comprises of piston, piston ring and cylinder liner materials.

Material selection for piston ring assembly is very essential to know the

characteristic of the material to assist in the design and for the success of the

operation.

3.1.1 Techniques of Material Selection

Most modern compressors use polymer based material for non-lubricated

piston rings although there are many older design compressors still in service that use

the bronze material. In design for a piston ring, the material selection is important

aspect to study to get the optimal design. The flowchart to illustrate a knowledge-

based material selection is as shown in Figure 3.1.

The main criterion of selection is the maximum acceptable-level of friction

and wears that can be tolerated by material during the sliding contact. In addition to

satisfy tribological requirements, the piston ring material must support the applied

33

load without significant distortion under the operating conditions, practically to

prevent severe surface damage during sliding. Five groups of materials can be

identified for having such an exhibit relatively lower friction and wear characteristics

under the non-lubricated sliding conditions:

• Polymers and polymer composites

• Solid lubricants

• Self-lubricating bronzes

• Carbons and graphite

• Hard-facing alloys, ceramics and cermets

Start

Component Category

Shape Category

Operational Factors

Process Characteristics Material Characteristics

Selected Materials and Processes

Materials Data and Optimization

Figure 3.1 Flowchart of material selection

34

The difference of mechanical, thermal and tribological properties of the

materials satisfies a wide range of the compressor’s operating specifications.

Moreover they included values of material capability to carry out continuous

operation in extremes of temperature, from -250°C to in excess of 500°C, at

pressures in excess of 1000 bar and at sliding speeds over 10 m/s.

Theoretically, there are at least three different techniques to obtain the

optimum material may be selected: (Smith, 1994)

1. The ‘Classical Procedure‘ using functional analysis and property

specification.

2. The ‘Imitative Procedure’ which consists the finding on what and which

material has been used for a similar component.

3. The ‘Comparative Procedure’ which consists of some postulates that the

component could be made firstly from some cheap and well-understood

engineering material, then assessing some ways such of each material’s

performance that would be possibly inadequate and from this step arriving

progressively to the right material.

The Classical Procedure is the only one that being universally applicable and

it is essential, even when procedures 2nd or 3rd are followed to check the findings by

functional analysis and property specification. However, the Classical Procedure is

expensive, time consuming and requires a considerable amount of prototype testing

to ensure that no critical requirement or essential property has been over-looked.

The Imitative and Comparative Procedures, if it’s applicable will provide

invaluable shortcuts, except for a greater expenditure of time and money, moreover

it will help to ensure that no essential parameter has been over-looked. The optimum

material for Piston Ring assembly (PRA), the imitative procedure was conducted in

this study. The reverse engineering was also carefully observed and investigated with

similar conditions, especially the case of the Scotch Yoke gas compressor with 245

bar discharge pressure. On the other hand, besides material selection the literatures of

the four stages and oil-free lubrication were studied in the same time and will be

compared at this study.

35

3.1.2 Piston Ring Design and Material

Based on the discussion in chapter II, the non-lubricated piston ring material

can be applied for gas compressor to ensure the purity of the gas during compression

process. Durability of non-lubricated compressor depends on anti-wear ability of

piston rings, as well as on the sufficient strength of the ring required at a high

pressure condition. To withstand the high pressure of 250 bar, several piston rings

are needed in order to reduce the load and friction per ring.

Conducting such an imitative procedure technique, the reverse engineering

was conducted and studied. Nishikawa (1998) reported the detailed development of

non-lubricated piston ring material for Scotch Yoke compressor with 245 bar

discharge pressure, 3m3/hr of flow rate, wide operating temperature range (from -

40°C to +40°C), easy service and long service intervals. In experimental test,

Nishikawa has chosen the polymer (high temperature plastic) as piston ring material

such as pure polytetrafluoroethelene (PTFE), pure polyetheretherketone (PEEK) and

polyimide (PI), and they have been tested by ring on disk wear machine test. During

the testing, the counter-face for cylinder material was from anodic oxide coating

aluminium. The pure PTFE ring material was not sufficient for Scotch Yoke

compressor, so that it needed improvement. To improve the PTFE, another filler was

added to basin resin PTFE, as shown in Figure 3.2. Unfortunately, Nishikawa did not

explain the material properties and the filler that added PTFE.

PTFE piston

Figure 3.2 Scotch Yoke compressor and PTFE added filler as piston ring

36

However, this reverse engineering exercise provided the helpful guidance on

determining the basic material that suitable for higher pressure gas compressor this

exercise also gave ideas for solving some specific problems that possibly could arise

for similar type of application. Therefore, a successful test on a polymer based

material (PTFE and PEEK) was conducted and this material is selected for use as the

piston ring for the new HRA symmetrical wobble plate compressor. Another reason

for the selection is that the raw materials and the fillers are available locally at a

reasonable price.

3.1.3 Piston Design and Material

Commonly, the design and materials used for compressor pistons will vary

with the make, type, and application of the compressor. They are designed to take

into account a number of conditions:

• Cylinder bore diameter

• Discharge pressure

• Rotational speed

• Stroke

• Required piston weight

• Strength, for differential pressure and temperature

Compressor pistons are typically designed as one of three types (Figure 3.3);

One piece, either solid cast iron or steel, for small bores and high pressure

differential applications, or one piece hollow-cored cast iron or aluminum, for large

diameter and lower pressures. Two pieces, aluminum or cast iron, which is split for

ease of hollow casting and weight control and generally used above 10-inch bore

diameters. Aluminum is used when the reciprocating weight must be reduced. Three

pieces, in which a ring carrier is added to permit band-type rider rings to be installed

37

directly into the piston grooves. While this design adds a part, it allows thicker rings

to be used since the ring does not has to be stretched over the outside diameter during

assembly. It is also used as a carrier for the piston rings on large diameter pistons.

Figure 3.3 Typical piston designs

(Bloch and Hoefner, 1996)

Materials commonly used for compressor pistons are aluminum, cast iron,

and steel. Aluminum is used when lightweight pistons are required in order to

balance reciprocating weights or to reduce inertia forces so they do not exceed rated

frame load limits. The aluminum used in common piston is a special alloy with a

tensile strength of 40,000 psi and a hardness of 100-110 Bhn. It may be given a

surface anodizing treatment to achieve a hardness of 370-475 Bhn, this improves

wear resistance. Applications are limited to approximately 200°F and a differential

pressure of 125 psi for castings (Bloch and Hoefner, 1996).

Cast Iron is the most common piston material for high strength and good

wear and corrosion resistance. It is used in either the cast or solid form, conforming

to ASTM A275. Steels are used for small bore, high differential pistons when

strength requirements are higher. They conform to ASTM A354 or A320. Steel is

also used in fabricating built-up type pistons in some designs.

However, in this new compressor, the high pressure reaches until 250 bars.

The steel material AISI 304 for piston was chosen to improve the strength of piston

38

and design in one piece. These piston have various diameter sizes; 36mm, 25mm,

15mm, and 10mm for 1st stage, 2nd stage, 3rd stage and 4th stage respectively. The

detail piston design is given in Appendix A and groove size for each piston and rider

rings is given in Appendix B.

3.1.4 Cylinder Liner Design and Material

The liner constitutes an important tribological element as a sliding surface

against the piston and piston rings. For non-lubricated material, the cylinder liner can

commonly be made of hard steel metal or hard-anodised aluminium (Baumann,

1994). The cylinder liner surface can be coated with a hard chromium layer to

improve the wear resistance of the cylinder liners (Affenzeller and Gläser, 1996).

The surface roughness also has significant effect on the tribological performance of

piston rings. This surface roughness affect is discussed in Chapter V.

3.2 Design and Analysis

The theory of piston rings has been applied to design and analysis of the

piston ring assembly in the proposed compressor. Principally, piston ring assembly in

linear motion has similar theory for all reciprocating engine. For a new symmetrical

wobble-plate compressor, the piston ring is considered ideal and followed by other

assumed parameters, as follows:

1. Highest loading condition, maximum gas compression force at each stage

was used which is at the Top Dead Centre (TDC) condition.

2. Piston side force and vibration effect are negligible hence piston slides in

linear motion.

39

3. Temperature during compression is stable and the cooling system is working

perfectly.

4. There is no gas leakage (blow-by)

5. The piston, piston ring, and cylinder liner are cylindricity (no out of

roundness).

3.2.1 Loads and Forces Acting On Piston Ring

In gasoline and diesel engine, the rings are generally of split-type

compression metal rings. When they are placed in the grooves of the piston, a

moving seal being formed between the piston and cylinder liner. Compression rings

normally of two or more pieces are located near the top of the piston, in order to

block the downward flow of gases from the compression chamber. For piston using

the lubricant, the oil rings must be placed below the compression rings to prevent or

control the passage of lubrication oil into the compression or combustion chamber.

In order to achieve efficient sealing, the piston ring should make a good fit

with both the cylinder liner and the top or bottom of the piston ring groove. The

radial fit is achieved if the inherent spring forces the ring, together with the pressure

of the working medium acting from behind the ring. In the case of a compression

engine, the working medium is the combustion gas. The gas pressure determines the

axial position of the ring within its groove, meanwhile the inertia and friction force

can alternate between the top and bottom of the groove as illustrated in Figure 3.4

(Society of Automotive Engineers (SAE), 1969).

40

Figure 3.4 Cross-sectional view of two pieces piston ring assembly in its sealing position (SAE, 1969).

Piston ring must provide an efficient sealing for the high pressure

compression gases. To achieve, they have primary sealing contact between face of

the ring and surface of cylinder liner, the secondary sealing contact is between the

side of the ring and ring groove, as shown by Figure 3.5. In addition to elastic ring

pressure, gas pressure behind the ring certainly affects the ring friction. Gas pressure

behind and between rings determines a function of cylinder pressure. Therefore,

contribution of the cylinder pressure to the ring friction is important during

compression process to ensure the sealing contact between ring and cylinder liner.

The piston ring should have a free shape in such a way that when it is inserted

in a cylinder it must provide a uniform pressure over the primary sealing contact.

Then the axial force needed for secondary sealing contact is mainly provided by the

cylinder pressure. From a friction point of view, having uniform pressure on the

primary contact is important. Therefore, determining and designing the free shape of

the ring to get such a uniform wall pressure when it is inserted is a challenging task

for this study.

41

Figure 3.5 Gas pressure on two pieces piston ring (SAE, 1969).

The ring is forced against the cylinder liner under a contact pressure that its

value depends on the dimension and total free gap of the ring and on the modulus of

elasticity of the material being used. The total free gap is defined as the distance,

measured along the neutral axis, between the ends of a piston ring in its

uncompressed state (Figure 3.6). A ring can be given a constant contact pressure, the

latter being a function of the angle.

The measurement of pressure is very difficult at inside compression cylinder.

In practice, it is therefore calculated from the tangential force. The tangential force is

the force applied tangential to the end of the ring and sufficient to close the ring into

the specified closed gap. By comparing the bending moment of the tangential force

against that of the constant contact pressure, the relationship of parameters on piston

ring is established.

3.2.2 Elastic of Piston Ring

In this study of piston ring it was necessary to estimate the elastic ring

pressure. The elastic ring pressure is defined as the pressure that the ring creates as a

result of its elastic deformation from the free to the loaded shape. Figure 3.6 is the

free body diagram of the ring. Symmetry about the x-axis is assumed and force P is

42

defined as the force which puts the ring into loaded shape. Strain energy of the ring

due to elastic deformation was calculated. Using Castigliano’s theorem, the relation

between force P and ring gap was derived. Later this force P is related to elastic ring

pressure. Referring to Figure 3.6, moment and force equilibrium equations give the

following set of equations:

(3.1) 0=− aPM z

0sincoscos1

=−− γγγ VPN (3.2)

0coscossin1

=+− γγγ VPN (3.3)

Figure 3.6 Free body diagram of a ring considering

elastic ring tension only

Moment arm a, was expressed as a function of mean ring radius and angle γ.

Equations 3.2 and 3.3 were solved for forces N and V. Equations 3.4, 3.5, 4.6 are the

expressions for bending moment, normal and shear forces respectively.

([ 1cos(1 )]γγ −−= mz rPM (3.4)

43

( )γγγγ sinsincoscos11

+= PN (3.5)

( )γγγγ sincoscossin11

+= PV (3.6)

Strain energy due to bending moment, normal force and shear force are given

by the following equations:

dxIE

MUzm

zb ∫=

2

21 (3.7)

dxAE

NUrm

n ∫=2

21 (3.8)

dxAG

vUr

s ∫=2

53 (3.9)

Castigliano’s theorem gives the deflections due to bending moment δb,

normal δn, and shear δs , forces. Following are the equations for deflections:

P

Ubb ∂

∂=δ (3.10)

P

Unn ∂

∂=δ (3.11)

P

U ss ∂

∂=δ (3.12)

The principal of superposition may by used to calculate total deflection δt:

snbt δδδδ ++= (3.13)

44

In terms of free, n and loaded q, ring gaps, total deflection is given by

Equation 3.14:

( qnt −=21δ ) (3.14)

Concentrated force P is not a function of x. This means that partial

differentiation with respect to force P can be performed before integration. It was

assumed that Ar, Em, G and Iz in equations 3.7 through 3.9 were constant and angle γ1

was zero. Integrations were performed with these assumptions. Equations 3.13 and

3.14 were used to derive the expression for force P.

( )

⎥⎦

⎤⎢⎣

⎡+

−

−=

z

m

rm

m

Ir

Ar

EqnP

23

1012172

2υπ (3.15)

For simplicity, elastic ring pressure was assumed to be uniform around the

ring. Figure 3.7 shows any elastic plane ring with an arbitrary shape loaded by a

uniform pressure P.

Figure 3.7 Plane elastic ring with uniform pressure

45

The moment about an arbitrary point D[x1,y1] is given by:

( )( )( ) ( )( )( )[ ]11 sincos yydstpxxdstpdM z −+−= λλ (3.16)

The differential ds may be expressed in terms of angle λ, dx and dy, and

equation 3.16 becomes:

( ) ( )[ dyyydxxxtpdM z 11 ]−+−= (3.17)

The resultant moment Mz about point D[x1,y1] may be obtained by integrating

equation 3.17 from B to C. This is a line integration and the right hand side of the

moment equation 3.17 is an exact differential. Hence the integration from B to C is

independent of the path. This result was applied to the piston ring. It was assumed

that the chord was approximately equal to the ring diameter.

Figure 3.8 Piston ring with uniform elastic ring pressure

Refering to Figure 3.8, the bending moment about point H is given by:

(3.18) 22 mm rtpMz =

46

Force P, which puts the ring into the same loaded shape, should create the

same bending moment:

(3.19) prMz m2=

Equation 4.18 and 4.19 were solved for the pressure pm:

m

m rtpp = (3.20)

It may be noted that the pressure given by the equation 3.20 is at mean ring

radius. Elastic ring pressure between the ring face and bore are proportional to ring

radius and is given by:

o

e rtpp = (3.21)

Equation 3.15 was used to calculate force p from ring geometry and material.

Elastic ring pressure was obtained from equation 3.21.

3.2.3 Piston Ring Forces

The forces that act on piston rings can be divided into two categories: axial

and radial forces. Axial forces determine the axial position of the piston ring at

different stages of the piston cycle. Radial forces determine whether a piston ring

collapses inward.

There are three primary forces that could influence the piston ring’s dynamics

include inertia, pressure and friction. The relative effects of theses forces on piston

rings are highly dependent on the operating conditions, such at certain high piston

speeds and low load, the inertia forces become more dominant than the pressure

47

forces. The opposite is also true; pressure forces are more dominant than inertial

forces at lower speeds and high loads. Frictional forces even at low speed and low

loads, do not have very much influences over the ring motion.

3.2.3.1 Inertial Force

The inertial force, Fl can be given simply as:

Fl = -m.a (3.22)

Where m is the mass of the ring and a is the axial acceleration of the ring. If

the ring is in contact with the piston groove, the ring’s axial acceleration is the same

as the piston. Pressure and friction forces would determine the ring’s acceleration if

the ring is floating in the piston groove.

3.2.3.2 Pressure Force

The pressure force, Fp acting on the bottom and top of the piston rings can be

modeled as;

Fp = P(r).A (3.23)

where P(r) is the pressure magnitude as a function of radius, and A is the

ring’s surface area. It was assumed that the pressure dropped linearly from the outer

to inner ring surface. The ring’s surface area A is;

A = π (D2 – d2) (3.24)

48

where D is the bore diameter and d varies depending upon the position of the

ring. If the ring is seated at the bottom of its groove, then d is the bore diameter

minus two times of the ring width. If the ring is touching the top of its groove, then d

is the piston diameter.

3.2.3.3 Friction Force

The friction forces, Ff between the piston’s ring face and cylinder liner can

be approximated by the semi-empirical equation (Mid –Michigan Research, 1998):

⎟⎟⎠

⎞⎜⎜⎝

⎛+=

pFDtpFf ax

axµν

τπ 8.4))(( (3.25)

Where p is the pressure behind the ring, tax is the axial width of the ring, Fτ is

the force due to ring tension, µ is the kinematics coefficient of friction and υax is the

ring’s axial velocity.

3.3 Ring Flutter

This "fluttering" motion occurs because of the pressure and inertial forces

continually alternate dominantly as the gas-flow’s paths change due to the ring

motion. Although the relative magnitude of the friction force is much less than either

the pressure forces or inertial forces, it becomes an important factor when the

pressure and inertial forces are nearly equal.

When pressure and inertial forces on a piston ring are close in magnitude, the

piston ring will have a tendency to oscillate axially faster in its groove. In the

49

instance in which ring flutter occurs near the end of the compression stroke, inertial

and pressure forces move the ring to the top of its groove. If the ring moves to the top

of its groove, it will floats and thus a flow path is opened between the ring bottom

and the groove bottom. As the ring reaches the top of its groove, the increasing

pressure forces at the end of the compression stroke can push the ring back down.

Before the ring is properly seated at the bottom of its groove. Gas can pass-through

behind the floating ring's groove and equalizing the pressure both the above and

below the ring. The inertial force then causes the ring to return again to the top of the

groove. This motion then can be repeated itself, impairing the ring's sealing

objective.

Friction actually acts beneficially during the presence of conditions favorable

to ring flutter. Since friction forces oppose any ring motion, they also dampen these

ring flutter oscillations and help to inhibit conditions that would initiate the

phenomenon. Although friction causes the piston ring wear and cylinder liner wear,

its very limited presence often improves ring stability. In some instances, ring flutter

can be a contributing factor to another unwanted ring behavior of the so-called ring

collapse. Further, ring flutter is sometime the practical result of ring collapse and is

to be an indication that ring collapse is likely to be identified. Ring collapse will be

explained in detail in the next section.

3.4 Ring Collapse

Ring collapse is an undesirable phenomenon in which the piston ring moves

radially. This is when the ring face being separated from the cylinder wall and

creating a large area in which compression gases can easily flow past. Ring collapse

is much more damaging to slower down performance than ring flutter. This condition

occurs when the forces acting on the ring face exceeded the combined forces

between the ring tension and the gas pressure behind the ring. Conditions favorable

for ring collapse usually coincide with high engine speeds and low engine load.

50

These operating conditions often cause a piston ring to seat at the top of its piston

groove near the end of a compression stroke; the high inertial forces (due to the high

engine speed) combined with a relatively low pressure-forces (due to the load engine

load) would keep the ring from seating on the bottom of its groove.

It is undesirable to have the piston ring seated at the top of its groove, since

that means the pressure gases below the ring (which are lower in magnitude than the

pressure of the gases above the ring) are supporting the ring radially. Thus, the

pressure force that applied to the ring face could overcome the ring tension and gas

pressure exerted from behind the ring. A force balance diagram for a scenario

favoring ring collapse is shown in Figure 3.9.

Figure 3.9 Force balance conditions for ring collapse

When a piston ring collapses, a large area is opened up between the ring face

and the cylinder liner. In order to find this area, the radial displacement of the piston

ring must be defined. The radial displacement at any point of a ring is given by:

⎟⎠⎞

⎜⎝⎛ +−= θθθθδ sin

2cos1)(

4

EIPR

r (3.26)

Where θ is the angular location of a point on a ring, P is the net radial load

due to pressure and ring tension, R is the central radius of the ring, E is the modulus

of elasticity of the ring material, and I is the moment of inertia of the ring cross

section. Zero radians for θ are located on the side of the ring opposite of the ring gap.

The additional gas flow path that opens up in the event of ring collapse can then be

defined as the area added to the ring gap area in the gas flow calculations:

51

(3.27) ∫=π

θδ0

2 RdA r

Substituting Equation (3.26) into Equation (3.27), and solving the integral yields:

EIPRA

53π= (3.28)

Under the circumstance that the ring end gap clearance is small, it is possible

that the radial displacement may be restricted due to the ends of the ring butting. To

find the maximum flow area that can occur between the ring face and the cylinder

liner, consider the area occupied by the piston ring, if its end clearance is zero.

Before finding the area, the effective ring parimeter, Peff could be calculated by using

equation:

endgapeff xDP −= π (3.29)

where D is the bore diameter (the original diameter of the outer ring surface)

and xendgap is the end gap clearance. An effective ring outer diameter Deff could be

derived from Equation (3.29) as:

π

endgapeff

xDD −= (3.30)

The area occupied by the ring, Aring is simply:

4

)( 2eff

ring

DA π= (3.31)

The maximum flow area, Amax is by definition to be:

(3.32) ringbore AAA −=max

52

where, 4

2DAboreπ

=

So Equation (3.32) can be simplified in term of end gap clearance and bore diameter

as:

DxA endgap21

max = (3.33)

3.5 Gas Leakage on Piston Rings

The high pressure variance in cylinder has great significance to decide gas

leakage. However, some past workers when calculating gas leakage often take the

pressure in cylinder as average value within a cycle or use the pressure variance in a

theoretical cycle as original data that makes it impossible to calculate the

instantaneous pressure distribution between rings and gas leakage through each ring.

Although sometimes the actual pressure in a cylinder is used, the data frequently

comes from measured value, and the use of some mathematical equations to predict

the leakage on piston ring is important in knowing the sealing characteristics of

piston rings and to carry out an optimal design (Yong et. al, 1986).

Predicting gas leakage during operation is very complicated and difficult to

measure. Theoretically, estimations may be used for some of the solutions, such as in

an oil-free gas compressor where there are three possible paths of gas leakage

through the rings as Figure 3.10 shows: (a) Between the rings and the surface of the

cylinder liner, (b) Between the rings and the groove of the piston, and (c) Through

the gaps of rings. For the piston with several rings, the total leakage mass flow ( m& )

through the ring is:

cibiaii mmmm &&&& ++= (3.34)

53

c

a

b

c b

a

Figure 3.10 Three possible paths of gas leakage through the rings

The flow through the gaps of the piston rings is presumed as one dimensional

compressible isentropic flow. Therefore the gas leakage through the gap of each ring

would be written as:

kk

i

ii

i

iici P

PTT

PAm

1

1

1

1 1[.−

+

+

+⎟⎟⎠

⎞⎜⎜⎝

⎛−=& (3.35)

where, )1(

2−

=kR

kfA ii α (3.36)

But if the flow speed in the gap equal to the speed of sound,

i

iici T

PBm =& (3.37)

where, ⎟⎠⎞

⎜⎝⎛

−⎟⎠⎞

⎜⎝⎛

++=

12

12

)1(2

kkkRkfB ii α (3.38)

54

The flows between the piston rings and the cylinder liner and between the

piston rings and the piston groove are considered as the flow in a thin clearance

between two smooth surfaces. This problem can be solved by using two dimensional

incompressible viscous laminar flow theories. Navier-Stokes equation was used to

predict the leakage between the rings and the cylinder liner ( ), and the leakage

between the rings and the piston groove ( ):

aim&

bim&

hTRppDm

i

iiaai µ

π24

)( 21

23+−∂

=& (3.39)

⎟⎠⎞

⎜⎝⎛

−

−∂= +

bDDTR

ppm

i

iibbi

2ln12

)( 21

23

µ

π& (3.40)

now, h

DC24π

= (3.41)

and, ⎟⎠⎞

⎜⎝⎛

−

=

bDD

E

2ln12

π (3.42)

The total leakage through the ring is thus:

i

iibak

k

i

ii

i

iii TR

ppECP

PT

TPA

mµ

)(1[

. 21

2331

1

1

1 +

−

+

+

+ −∂+∂+⎟⎟

⎠

⎞⎜⎜⎝

⎛−=& (3.43)

This formula which predicts a flow leakage through piston rings is usually

used for the butt joint type of piston rings. Several types of cut joints were designed

to minimise leakage. Scarf joint and step-cut joint were a more popular shapes for the

use in high-pressure oil-free compressors. These joints are either more easily cut,

installed and removed than compared with the gastight joint construction that costs

up to twice as much. In accordance with the sealing function required, piston ring

joints can be shaped to any degree of complexity within the bounds determined by

the material properties and ring dimension. To minimise the leakage between piston

55

ring cylinder liners, backup springs could be attached behind the actual ring to give a

slight amount of pressure (1-3 bar) and assist the piston ring in establishing the initial

seal.

Butt Joint Scarf Joint Step-cut Joint

Gastight Joint Two-piece Joint

Figure 3.11 Various types of cut joints for piston rings

3.6 Wear of Piston Ring – Cylinder Liner

Wear defined as the removal of material from solid surfaces to be a result of

mechanical action. Wear due to sliding is usually a very slow process, it is very

steady and continuous. From the reciprocating motion of the piston ring against the

cylinder liner causes the wearing over the contact surface, as shown in Figure 5.1.

Generally, wear is caused by individual and combined effects among corrosion,

adhesion and abrasion (Chui, 2001). However, corrosive wear normally occurs in

cylinder of a combustion engine which produces the harsh exhaust gases. It is quite

reasonable to assume corrosive wear as negligible in a gas compressor.

56

Figure 3.12 Rubbing of two contact surface under microscopic view (Chui, 2001)

Already mentioned earlier the abrasive wear occurs when there are impurities

present at the contact surfaces. The impurities can come from atmospheric dust and

from the solid debris from corrosive and adhesive wear. When the impurities stay

between two contact surfaces, such as the piston ring and cylinder liner, the ratio of

asperity contact is increased.

Adhesive wear mainly affects parts of the piston ring where there is solid-to-

solid surface sliding contact between piston-piston ring-cylinder liners. The removal

of material takes a form of small particles which are usually transferred to the other

surface. For automotive engine, the adhesive wear is especially significant when the

engine was at the beginning of operation and was cold, due to insufficient oil in the

piston ring-cylinder liner. For non-lubricated piston ring, the adhesive wear is also

very significant because no oil as lubrication to decrease the debris of material. This

type of wear is the focus of discussion in this thesis.

57

3.6.1 Surface Texture of Piston Ring

Surface texture affects the wear rate of the surface of piston ring-cylinder

liner. Surface texture can be identified with the roughness and smoothness of a

surface. A complete description of a surface involves many aspects; the amplitude

(roughness and smoothness), the waviness (the pattern of surface roughness) and

density of surface asperities per unit surface area.

There are several methods to measure of the surface texture. A common

parameter detecting a surface texture is Ten-point-height (Rz). It is the average

distance between the five highest peaks and the five deepest valleys of a surface

within the sampling length. The formula for Rz is:

R =5

)54321()54321( YvYVYvYvYvYpYpYpYpYp ++++−++++ (3.44)

Where: Yp = The highest of the five highest peaks

Yv = The depth of the five lowest valleys

Another commonly used parameter is the Roughness Average, (Ra).

Roughness average is the arithmetic average of the distance of the filtered or

unfiltered roughness profile from its mean line, as shown in Figure 3.13:

Figure 3.13 Surface profile (Chui, 2001)

58

The formula for roughness average is as follows:

Ra = L1 dxxh

L

∫0 )( (3.45)

Another parameter is the Root Mean Square Roughness Average (Rq). It is

the root mean square of the distance of the filtered or unfiltered roughness profile

from its mean line, as shown in Figure 3.13. Rq is the standard deviation of the

amplitude density distribution. The amplitude of this parameter is more sensitive to

the value of its peak and valley compared to Ra. The formula for root mean square

roughness average is as follows:

Rq = L1 (3.46) ∫

L

dxxh0

2 )(

Using Ra and Rq to measure a surface texture is sufficient, except when there

are differences in the density of the real contact area. To illustrate this point, consider

two surfaces that have the same values of Ra and Rq but have different texture.

Figure 3.14 shows the two different surfaces, which are between surface 1 and

surface 2.

Figure 3.14 Two different surface profiles (Chui, 2001)

59

Applying Equation 3.45 and equation 3.46 will give same value of Ra and

Rq, even though these two surfaces are different. To distinguish surface 1 from

surface 2, another parameter is introduced: the Abbott Firestone Curve (AFC), which

is also called the Bearing Ratio Curve. It is a graphical representation of the bearing

ratio parameter, tp (length of bearing surface, expressed as a percentage of the

assessment length of a surface specimen, at a depth below the highest peak or a

selected distance from the average reference) in relation to the profile level. The

curve contains all the amplitude information of the surface profile. The AFC for

surface 1 and 2 is shown in Figure 3.15.

Figure 3.15 Surface representation using Abbott Firestone Curve (AFC)

(Chui, 2001)

From the Abbott Firestone Curve (AFC), the simple difference between the

two surfaces can be identified where surface 1 has lower amplitude density than

surface2.

The surface of the cylinder liner needs to be measured too. However, this step

can be avoided by assuming that the interaction of two surfaces can be simplified as

the interaction of one rough surface with a planar surface. The roughness value will

become the combination of two surface roughnesses.

60

Rz = Rz1 + Rz2 (3.47)

Where: Rz = the roughness of the combined surface profile

Rz1 = the roughness of surface 1

Rz2 = the roughness of surface 2

3.6.2 Wear Mechanism

Considering a small part of piston ring sliding on a cylinder liner as shown in

Figure 3.16 then this part of the piston ring can be assumed to be very small since the

curvature effect of the piston ring in a circumferential direction is negligible. The

amount of worn volume from the piston ring surface is calculated using the following

equation:

= V∂ sHFk ∂ (3.48)

Where:

V∂ = Incremental volume loss from contact surfaces

= Sliding distance of one surface with respect to another surface s∂

F = The total load applied to the piston ring surface while it is sliding

H = Combined hardness of the piston ring and cylinder liner

k = Wear coefficient

61

Figure 3.16 Small section of piston ring sliding on cylinder liner

To evaluate the total load that applies on the surface, it can consider that a

piston ring always pressed against the cylinder bore by forces such as ring tension

and the pressure forces behind the piston ring. Assuming that all the forces are

transferred into the cylinder liner as a normal force and by using a solid mechanic

analysis, then the value of the pressure on the ring against cylinder bore will result in

axial force distribution on the piston ring surface, as shown in Figure 3.17.

Figure 3.17 Ring section contact with cylinder liner

62

Applying Newton Second Law, one can deduce that:

(3.49) rFN =

where

(3.50) ∫= dxxfFr )(

In order to obtain f(x), the following assumptions are made:

• Linear elastic behavior of the piston ring material

• Modulus of elasticity of the cylinder liner is infinite

• Deformation is very small

• Shear stress on the piston ring is negligible compared to normal stress

With the assumptions above, the determination of force distribution of the

piston ring surface into the slices partition with the same width can be obtained as

illustrated in Figure 3.18. These slices are assumed to be completely separated from

each other since the normal stress is much higher than the shear stress.

Figure 3.18 Piston rings in slices partition

63

Using Hook’s law:

ii Eεσ = (3.51)

where: i

i

ll∆

=ε (3.52)

ii Afi σ= (3.53)

In order to obtain the forces for each slice, an arbitrary reference point is

assigned to start the process of evaluating the total force. The process continues with

the reference point being varied until the criterion of equation 3.49 is met. The final

value for the force of each slice will be used for the calculation of wear at each slice.

3.7 Symmetrical Multistage Wobble-plate Compressor

The new symmetrical multi-stage wobble-plate compressor has been

developed as a slow Home Refueling Appliance (HRA) and the prototype has been

patented in Malaysia under the title “Wobble Plate Compressor” with filing number,

PI 2005 5456 (Musa, 2005), Appendix D. The compressor design requirements for

this new compressor design are given in Table 3.1:

Table 3.1 Home Refueling Appliance (HRA) compressor design requirement:

Design requirement Value Inlet pressure 50 Psig (3.5Bar) Outlet pressure 3600 Psig (250 Bar) Gas flow-rate 1 m3/hr (slow refueling) Working gas Natural gas (gas compression index = 1.27) Operating speed 1000 rpm Overall Light, small, compact, low vibration and noise Lubrication method Oil-free cylinder wall

64

The initial symmetrical multi-stage wobble-plate compressor design

developed for the natural gas Home Refueling Appliance (HRA) is illustrated in

Figure 3.19. The main design features that differentiate this new symmetrical multi-

stage wobble-plate compressor from the existing wobble-plate compressor are the

use of the multi-size piston diameter for the multi-stage compression, the use of the

end-joint pairs as the connecting rod, forming symmetrical wobble-plate and piston

arrangement at each compressor ends and the use of oil-free lubrication system

inside cylinder assembly.

Those differences determined the unique mechanism of the compressor when

it was operated. The wobble-plate bearing has the functions as the sliding interface

between rotor and wobble-plate. The inner reel of wobble-plate bearing was tightly

fitted with rotor while the outer reel of the bearing was tightly fitted with the wobble-

plate. Through the assembly among the shaft, rotor, wobble-plate bearing and

wobble-plate, then each shaft rotation induced the wobble-plate wobbling motion.

Afterwards the wobbling motion was transferred to be the movement of piston

reciprocating via the certain connecting rod, which connects the piston and the

wobble-plate with a ball-joint connection between both ends of the connecting rod.

To prevent the connecting rod from being tangled together, the wobble-plate was

constrained from having rotated along with the rotor by using a single anti-rotation

mechanism which comprises the function of the anti-rotation ball, the anti-rotation

shoe and the anti-rotation rod.

Focusing on the gas flowing inside the new compressor, the reciprocating

motion of piston firstly compressed gas in the cylinder block, then gas was

transferred between each stage by using inter-stage piping. To achieve a higher

pressure, the multi-stage compression was necessary. Therefore, four stage

compressions were used together with four different sizes of piston, cylinder liner

and cylinder block. Fixed capacity configuration of the part was used for this new

compressor design, due to the fixed flow-rate requirement being required as it is

given in Table 3.2. In the following sub-sections, further details of initial components

for piston-piston and ring-liner design in this compressor will be outlined and its

design considerations will be explained.

65

Wobble-plate pin Wobble-plate bearing End plate bearing Piston Shaft

r Valve

Line

Figure 3.19 Cross section of the symmetrical multi-stage wobble-plate compressor

More details on the typical piston assembly in this compressor design are

shown in Figure 3.20. The piston design was made different from that of the existing

piston used in an automotive air-conditioning using refrigerant compressor which

also uses wobble-plate compressor concept. The piston for higher pressure

compression tended to be more solid in design. The first-stage of the piston diameter

size was made as the largest one followed subsequently by the next stages and the

smallest piston diameter size is for the last stage. However, the number of piston ring

required for sealing at the piston for each stage would be different depending on the

increasing value of the compression pressure (Chlumsky, 1965).

Rotor

Wobble-plate

Anti-rotation ball

Anti-rotation ball

Shaft pin

Cylinder block

66

Connecting rod Piston pin Wobble-plate

pin Piston coupler

Bush PistonBush

Figure 3.20 Cross-section of piston assembly (First stage piston)

For connection among the parts throughout this compressor design, a pair of

self-lubricated end-joints, male and female were fastened altogether to make a single

connecting rod. Thread engagement length between the male and female end-joint

was adjusted according to the required connecting rod length. This threaded joint was

secured using nut and thread locker sealant. Bushes were used to position the end-

joint balls at the piston couplers and at the wobble-plate.

3.8 Installation of Piston Ring

Installations of piston ring into the groove were a tricky challenge during

assembly. Extreme difficulty was experienced when the un-cut piston ring needed to

be installed in a groove of a rigid piston. Usually the installation methods have used

either the split piston or cut the piston ring. But some disadvantages were found since

a split piston requires exact precision in its manufacture. The cut piston ring was

easily installed into the groove but the possible gas leakage (from blow-by) from the

gap of the cut which was very critical. Another method was to heat the piston ring in

a hot oil bath with 80˚C-100˚C of temperature. After the piston ring become soften, it

was easy to expand the ring into the groove and clamped until the material returning

to the original form. Yet this process was impractical as it changed the properties of

the material with respect to strength and therefore was not adopted.

Male end-joint

Female end-joint

End-joint ball

End-joint ball

Wobble-plate

67

In this study, another method was proposed, it is by using three pieces of

installation tools; installation sleeve, expanding sleeve, and sizing sleeve. All these

tools were made from a polymer material with good sliding characteristics and low

abrasiveness to avoid damage to the piston ring. The illustrated procedures are shown

at Figure 3.21. With the o-ring in position, the piston ring was inserted into the

installation sleeve as far as possible and then it was pushed further in by using the

expanding sleeve over the installation sleeve until the ring is rested in the groove

around the o-ring. The sizing sleeve is used to check whether the ring will perform

its function as intended. The following points should be observed prior to the

installation of the seals:

• Ensure the piston groove has a lead in chamfer; if not, use an installation

sleeve

• Remove machining residues from such chips, dirt and other particles and

carefully clean all parts

• The piston ring can be installed much easily if they were greased or oiled.

• Use no sharp-edged installation tools

68

O-ring Un-cut piston seal

Installation sleeve

Piston groove Step 1:

Push

Expanding sleeve

Step 2:

Sizing the piston ring using the sizing sleeve Step 3:

Figure 3.21 Installation processes of the un-cut piston ring into the piston groove

CHAPTER IV

EXPERIMENTAL METHODOLOGY AND PROCEDURES

4.1 Introduction

This chapter describes the experimental methodology and procedure to

accommodate the objective stipulated in Chapter I. It consists of two sections.

Section 4.2 describes the experimental of material for piston ring assembly using the

test bench to show the performance and influence the tribological effects of materials

such as surface roughness, coefficient of friction, and wear rate. Section 4.3,

describes all the experiments and procedures of prototype of compressor using

selected material of piston ring assembly.

4.2 Tribotest

Traditional way of measuring ring-liner wear in an existing reciprocating

compressor is the gauge method, which measures the increase of the bore diameter as

a result of wear. To measure the wear using the gauge method, the compressor

cylinder needs to be dismantled. As such ring-liner wear tests are expensive and time

consuming. To reduce time and cost, many attempts have been made to estimate

70

ring-liner wear using the test benches. Rotary methods include pin on disk (ASTM

G99) and the block on ring (ASTM G77) is a common wear standard test bench.

Other test bench devices include the ring on ring, ball on flat, four balls, and thrust

washer methods. Prior to discussion on how this wear test is conducted this chapter

touches first on the friction and wear of piston rings assembly and how these factors

influence the tribological effects.

As discussed previously, the friction and wear of piston ring-cylinder liner is

important to be predicted so as to know the tribological effect. Some of principal

advantages of using the standard wear test methods are (Blau, 1999):

• The previous test methods have been carefully evaluated and

documented.

• The repeatability and reproducibility of results tends to be better

documented and understood than certain tests for particular

specialisation or one of other types of wear testing machines.

• In many cases, a great deal of previous data exist and it is convenient

to compare new results with the existing data.

• Documentation and reporting requirement have been established so

that all the major variables and results of the work can be presented in

a complete and organised manner.

In this study, the reciprocating sliding standard wear test (ASTM G-132) was

selected to predict the wear, the coefficient of friction, normal load and frictional

force for a piston ring and cylinder liner. The brand of machine is DUCOM

Reciprocatory Friction and Wear Test Monitoring Machine (Figure 4.1). This

machine also includes the heater up to 150 deg Celsius, humidity sensor, and variant

weight load 10 N to 100 N.

71

Monitor System TR-281m8

Wear Test Machine

Humidity System

Figure 4.1 DUCOM Reciprocatory Friction and Wear Test Monitoring Machine

4.2.1 Material and Specimen Preparation

The reciprocatory sliding test used two samples in the form of a pin for piston

ring specimen and a plate as cylinder liner specimen. The dimension for pin was 3

mm in radius and 6 mm in length. For the plate, the dimension was 15 mm length, 15

mm width, and 2 mm thickness. The material as having discussed in Chapter III, the

material of pin to be a piston ring is PTFE and PEEK. For material of plate for

cylinder liner is XW41 hard chrome coated.

From the literature review, it is generally accepted that the cylinder liner

should be much harder than the polymer and hardened steel, as it was often

recommended. Cylinder liner in common practice also has the smooth surface

roughness. The cylinder liner roughness has a rather more complex effect on polymer

wear. While it has often been suggested that the roughness of material should be as

low as possible to reduce abrasion of the polymer. In order to get the lower wear rate,

72

the cylinder liner was chosen to be harder than piston ring and smoother in surface

with the surface roughness below than 0.4 µm (Zhang et. al, 2005). Some

experiments were measuring surface roughness of plate as cylinder liner by using the

standard MITOTUYO Surface Roughness Machine. The result of average surface

roughness of this test was about 0.124 µm and maximum 0.142 µm, as being shown

in Figure 4.2.

Figure 4.2 Surface roughness tested for cylinder liner

(Material XW41 hard chrome coated)

4.2.2 Experimental Procedures

Friction and wear test were performed by using a pin on a plate arrangement

on a DUCOM Reciprocatory Friction and Wear Test Monitoring Machine. The plate

was fixed on the pad and the pin was attached on a bracket of the reciprocating arm.

The tests were conducted using a beam at various load applied ranging from 20N to

50N. The heater was controlled at 100˚C to give a testing temperature; the stroke

length was about 10 mm; sliding velocity 4 m/s; and test duration per sample was

73

one hour. The humidity environment was set at RH 50-60%. All of the tests were

done and monitored by sensors that being built-in to the testing machine.

After the test, the pin was removed and cleaned from any debris. The pin was

weighed by using OHAUS digital precision balance (four decimals) and their values

which also listed the weight loss, i.e. the differences in specimen weight, before the

test and after the test. This experimental procedure was repeated for PTFE and PEEK

materials each with different plate.

4.3 Prototype Experimental Test

This section discusses the development of the prototype, the development of

the rig and procedures. The focus of the experiment was on the monitoring and the

evaluation on the performance of the designed piston rings. The test was to prove

that the piston ring performs well during compression of the air up to extremly high

pressure of 250 bar. The testing and commissioning of compressor and the rig was

carried out with some anxiety as to whether the analysis, the design and the material

selection were done correctly or otherwise. Only PTFE set of piston ring material has

been developed and tested. Therefore, the results presented in this chapter are those

values taken from experimental tests using air as the working fluid. Nevertheless, the

real tests on natural gas are to be organised by utilising all the experience and lesson

learnt from that on air. During testing and commissioning to the prototype

compressor and to the rig and this procedure took more than two months to complete

until the set-up was ready for the actual experimental work.

Testing was done in two modes, firstly was conducted without-load and

secondly being conducted with-load. In the without-load test, air at atmospheric

pressure entered the inlet section. Suction and discharge ports of each stage were left

open. This test was done to check the functionality and the kinematic of the motion

mechanism used to reciprocate the pistons inside the cylinder liners. Test without

load was essential to check the operation of the valves to ensure that the suction and

74

discharge valves work properly. The reading of the gauge pressure at every stage

gave indication whether suction or discharge valves were both closed, both opened or

either one was close or open.

4.3.1 Prototype and Rig Development

The first prototype symmetrical wobble-plate compressor for Home

Refuelling Appliance (HRA) has been developed, fabricated and tested (Figure 4.3).

To meet the HRA requirement, the compressor were designed compact, small, low

vibration, low noise, and low power consumption. Significant differences of features

which set this new symmetrical multi-stage wobble-plate compressor apart from the

existing single sided wobble-plate compressor were the use of the multi-size piston

diameter for the multi-stage compression, the use of end-joint pairs as the connecting

rod and the symmetry of the piston arrangement at each compressor ends. More

detailed specification of this compressor is given in Table 4.1.

Figure 4.3 First prototype of symmetrical wobble-plate compressor

Home Refuelling Appliance (HRA)

75

Table 4.1: Specification of symmetrical wobble plate compressor

Calculated Value Input Data First

Stage (P1)

Second Stage (P2)

Third Stage (P3)

Fourth Stage (P4)

Cylinder diameter 36 mm 25 mm 15mm 10 mm Suction Pressure 3.45 bar 10.04 bar 29.25 bar 85.21 bar

Discharge Pressure 10.04 bar 29.25 bar 85.21 bar 248 bar Fluid Air Stroke 10 mm

Mean speed 4 m/s Rotating Speed 1000 rpm Tilting Angle 5o

Pressure ratio 2.91 Capacity 1 m3/hr

The experimental rig for the prototype testing is shown in Figure 4.4. The rig

is composed of a motor to drive the shaft, storage tank, piping and the compressor

prototype. The instruments installed for the test are the flow-meter, the analogue

pressure gauges, the pressure sensors, the torque sensor and the thermocouples.

Compressor cooling was done by blowing air across the compressor casing as well as

across all eight cylinder blocks, using an external fan.

Pressure measurement for the right side of the symmetry and at the suction

port of the first stage cylinder only and the discharge ports of the all stages were

handled by the pressure sensors which are connected to a Data Acquisition System

(DAS). Whereas the same set of pressure measurement for the left side of the

symmetry were directly measured by the pressure gauges.

76

4.3.2 Experimental Procedures

The experiment was conducted according to the following procedures:

1. Check all measurement systems are in working order.

2. Ensure the compressor unit is in ready condition for test.

3. Set the pressure relief valve to about 250 bar.

4. Switch on the air supply auxiliary compressor as input fluid and adjust the

pressure of the air to about 3.5 bar. Which is the suction pressure of the

compressor under test.

5. Switch on the inverter and set the motor speed gradually to 1000 rpm.

6. Switch on the compressor motor and when condition has stabilised, record all

data by Data Acquisition System (DAS).

7. Finally, shut down the compressor power by gradually reducing the speed of

the compressor to 0 rpm.

Figure 4.4 Experimental Rig of HRA

CHAPTER V

RESULTS AND DISCUSSIONS

5.1 Introduction

This chapter presents the results and discussions of the experiments described

in Chapter IV. It consists of tribostest, Finite Element Method (FEM) and

experimental tests.

5.2 Tribotest

The main objectives of the experimental tests performed were to study the

trobological aspects of PTFE and PEEK as piston ring slides along a XW41 hard

chromed cylinder liner in various load conditions, as follows:

(i). The transfer film phenomenon (Figure 5.1).

(ii). The coefficient of friction

(iii). Wear

(iv). Weight losses

78

Plate as cylinder liner

Pin as piston ring material

The transfer film

Figure 5.1 Transfer film phenomenons

The tribotests on friction and wear have been done with various loading

values (Figure 5.2 – Figure 5.7). In this case, the temperature was set as independent

value in worst condition 100˚C, controlled with humidity system. Usually the

compressor was designed to have the intercooler for each stage. The intercooler will

convert the high temperature to ambient temperature around 30˚C - 35˚C. However

there was not significant increase in temperature at each stage of compressor. Hence

the stage intercooler was never used.

79

Figure 5.2 Coefficient of friction of PTFE vs XW41 hard chrome coated

Figure 5.3 Coefficient of friction of PEEK vs XW41 hard chrome coated

80

Figure 5.4 Wear of PTFE vs XW41 hard chrome coated

Figure 5.5 Wear of PEEK vs XW41 hard chrome coated

81

Figure 5.6 Temperature of PEEK vs XW41 hard chrome coated

Figure 5.7 Humidity of PEEK vs XW41 hard chrome coated

82

Table 5.1 Weight loss of PTFE material vs XW41 hard chrome coated: (gram)

Load Before After (N) Plate Pin Plate Pin

Weigh Loss of Pin(gr)

Weigh Loss of

Plate(gr) 20 6.9715 0.5277 6.9714 0.5245 0.0032 0.0001 30 7.259 0.5104 7.2588 0.4908 0.0096 0.0001 40 7.1765 0.5223 7.1763 0.5034 0.0189 0.0002 50 7.3704 0.5219 7.3703 0.4904 0.0215 0.0001

Table 5.2 Weight loss of PEEK material vs XW41 hard chrome coated: (gram)

Load Before After (N) Plate Pin Plate Pin

Weigh Loss of Pin(gr)

Weigh Loss of

Plate(gr) 20 6.9717 0.2055 6.9717 0.2148 0.0093 0 30 7.2407 0.211 7.2407 0.2015 0.0099 0 40 7.1796 0.2225 7.1795 0.2169 0.012 0.0001 50 6.9017 0.218 6.9017 0.2083 0.022 0

Weight Loss of Pin (PTFE)

0

0.005

0.01

0.015

0.02

0.025

20 30 40 50

Load (Newton)

Wei

ght L

oss

(Gra

m)

Figure 5.8 Weight loss of pin (PTFE)

83

Weight Loss of Plate (PTFE)

0

0.00005

0.0001

0.00015

0.0002

0.00025

20 30 40 50

Load (Newton)

Wei

ght L

oss

(Gra

m)

Figure 5.9 Weight loss of plate (PTFE)

Weight Loss of Pin (PEEK)

0

0.005

0.01

0.015

0.02

0.025

20 30 40 50

Load (Newton)

Wei

ght L

oss

(Gra

m)

Figure 5.10 Weight loss of pin (PEEK)

84

Weight Loss of Plate (PEEK)

0

0.0001

0.0002

20 30 40 50

Load (Newton)

Wei

ght L

oss

(Gra

m)

Figure 5.11 Weight loss of plate (PEEK)

5.2.1 Summary and Discussions

The DUCOM Reciprocatory Friction and Wear Test Monitoring Machine,

was a very effective tool for evaluating the tribological performance of a piston ring-

cylinder liner. Presented in Figure 5.2 and 5.3 are the friction coefficient of PTFE

and PEEK as a function of the sliding time of 1 hour under 20N, 30N, 40N, 50N

applied loads. It can be seen that the friction coefficients significantly follow a trend

graph where friction coefficient increases with respect to time sliding and arrives at a

relatively stable value at steady state condition. For PTFE, under 20N applied load

the friction coefficient have minimum value µ = 0.12. Under 30N, 40N, 50N applied

load, the friction coefficient showed a stable value ranging from µ = 0.20 to µ = 0.22.

This condition does not exceed much the steady state value of friction. These result

can be obtained as value of friction coefficient for PTFE sliding against XW41 hard

chromed. The same trend as showed for PEEK material where the steady state

friction coefficient were ranging from µ = 0.29 to µ = 0.37.

85

Variations of wear with sliding under load as a function of time are shown in

Figure 5.4 and Figure 5.5 respectively. For PTFE material, the differences of wear

under 20N and 50N load become more significant while time increased. However,

for PTFE and PEEK material, the wear rate shown the same trend and close value of

wear. In maximum load 50N, the wear of PTFE given the high wear rate W = 0.12

µm comparison with PEEK W = 0.10 µm. However the wear for 20N, 30N, 40N for

both materials are shown to be stable and in a steady state condition ranging from W

= 0.02 µm to W = 0.10 µm.

Overall, the results of weight loss of both materials are shown in Figure 5.8

and Figure 5.9 which show an agreement respectively with the established trends,

which showed that mass loss increased with the increasing of the normal load.

Significantly, higher mass losses of pin were under higher load 50 N around 0.02

grams. On the other hand, the cylinder liner specimen showed a negligible mass loss

under various load (Figure 5.10 and 5.11).

5.3 Design and Modeling Using Finite Element Analysis (FEA)

Static analyses were done on the critical parts which are the piston, cylinder

liner, and piston rings. These parameters are as an input data from analysis in section

above are used to calculate the load acting on each part of Piston Ring Assembly

(PRA). In the analysis, the gas compression force during TDC for each stage was

used as the loading input. Through this analysis, stresses on the Piston Ring

Assembly (PRA) parts and its strength can be assessed.

86

Table 5.3 Results for stress analysis on the pistons:

Stage Yield strength (MPa)

Maximum Von mises stress (MPa)

Maximum displacement (m)

Safety factor

1 215 19.47 2.42 x 10-6 11 2 215 13.50 1.81 x 10-6 15 3 215 12.13 1.61 x 10-6 17 4 215 34.64 4.41 x 10-6 6

Gas pressure on the piston

Fixed surfaces

Figure 5.12 Loads and boundary conditions of the piston (first stage)

Highest stress

Figure 5.13 Von Mises stress of the piston (first stage)

87

Figure 5.14 Deformation of the piston (first stage)

Figure 5.15 Factor of Safety of the piston (first stage)

88

Table 5.4 Results for stress analysis on the cylinder liner:

Stage Yield strength (MPa)

Maximum Von mises stress (MPa)

Maximum displacement (m)

Safety factor

1 215 2.75 7.42 x 10-8 75 2 215 6.44 1.74 x 10-7 32 3 215 17.68 4.05 x 10-7 12 4 215 53.61 1.24 x 10-6 4

Figure 5.16 Loads and boundary conditions of the cylinder liner (first stage)

Highest stress

Figure 5.17 Von Mises stress of the piston (first stage)

89

Figure 5.18 Deformation of the cylinder liner (first stage)

Figure 5.19 Factor of Safety of the cylinder liner (first stage)

90

Table 5.5 Results for stress analysis on the piston ring assembly (PTFE):

Stage Yield strength (MPa)

Maximum Von mises stress (MPa)

Maximum displacement (m)

Safety factor

1 28 2.55 1.45 x 10-7 11 2 28 4.02 2.19 x 10-6 7 3 28 8.23 4.87 x 10-6 3 4 28 15.56 5.25 x 10-6 1.8

Table 5.6 Results for stress analysis on the piston ring assembly (PEEK):

Stage Yield strength (MPa)

Maximum Von mises stress (MPa)

Maximum displacement (m)

Safety factor

1 110 3.95 2.23 x 10-7 28 2 110 6.48 3.58 x 10-6 17 3 110 13.76 6.54 x 10-6 8 4 110 36.68 7.86 x 10-6 3

Figure 5.20 Loads and boundary conditions of the piston ring (PTFE)

91

Figure 5.21 Von Mises stress of the piston ring (PTFE)

Figure 5.22 Deformation of the piston ring (PTFE)

92

Figure 5.23 Factor of Safety of the piston ring (PTFE)

5.3.1 Summary and Discussions on Piston

As mentioned earlier piston size for the first to fourth stage are 36 mm, 25

mm, 15 mm, and 10 mm. The geometry of piston was designed to ensure a minimum

misalignment and a smooth sliding in a linear motion during compression process

inside the cylinder liner. All pistons were machined from a one piece to obtain the

required shape. Each piston was made from a stainless steel AISI 304, with yield

strength of 215 Mpa.

In order to consider for the highest loading condition, maximum gas

compression force at each stage was used which is at the TDC condition. Bottom

surface of the piston was fixed. Gas pressure were transferred from top of piston

surface and being assumed that the gas went through to the first groove of the piston

ring. Due to simple symmetrical geometric shape, all pistons were modeled using

hexagonal element. Stress concentrations by compression occurred at the transition

cross section area at the piston. The results for the first piston are shown in Figure

5.12 to Figure 5.15. With the same procedures, for others piston are summary of

results given in Table 5.3. Overall, the static analyses for all pistons were done and

93

highest stress concentrations were seen at the neck between the piston rod and

coupler.

5.3.2 Summary and Discussions on Cylinder Liner

Analysis the first stage cylinder liner, stress concentrations occurred at the

location of the contact interaction between the rigid surfaces and the cylinder bore.

The outside diameter was fixed at its reference point. The maximum gas compression

force during Top Dead Centre (TDC) condition at each stage was inserted as a point

load at the bore of cylinder liner. At the top of cylinder liner, are the two bolt holes

meant for valve assembly. This region has high stress concentrations and

displacement but not effected to the cylinder liner as a part of compression process.

The same patent of stress concentration and displacement also occurred for the

analysis at the other stages. The results of the analysis for the other stages are given

in Table 5.4 and the FEA results for the first cylinder liner are shown in Figure 5.16

to Figure 5.19.

5.3.3 Summary and Discussions on Piston Rings

For piston ring analysis, the static analyses were done for both material,

PTFE and PEEK. Thus in this analysis, the multi-body contact analysis method (in 3-

Dimension) was used to obtain the boundary condition location where the piston ring

has interacted with piston and cylinder liner. In this case, the top piston ring is

critical part and worst condition because the gas compression was applied first at the

top of piston ring which is installed in the piston groove. Pressure loading from the

gas compression process was applied on the piston top surface. In actual condition,

the gas pressure was applied to the ring through the groove of the piston. At the TDC

and under ideal piston ring condition no gas leakage is assumed through the second

ring.

94

Each of the analysis models has different model geometry, material properties

and magnitude of loading but has the same boundary conditions and the results are as

summarised in Table 5.5 for PTFE and Table 5.6 for PEEK.

5.4 Prototype Experimental Result

As mentioned earlier the experiment was only ready to be conducted after the

compressor prototype and the rig have gone through rigorous test and

commissioning. Tests were started from a lower speed of around 100 rpm and being

increased until it reached 1000 rpm. The results of the experimental test are given in

Figures 5.24 to 5.27 which show the variation of the stage pressure versus operating

time with load.

Figure 5.24 Graph of pressure vs time, first test

95

Figure 5.25 Graph of pressure vs time, second test

Figure 5.26 Graph of pressure vs time, third test

96

Figure 5.27Graph of pressure vs time, fourth test

5.4.1 Summary and Discussions

The experiments carried out were actually a series of test and commissioning

of the prototype. These tests and commissioning was carried out for about one and

half years. Every failure experienced, the research went back to the drawing board

and to the machinist for any modification or rectifications job.

Prototype of the new symmetrical wobble-plate was tested successfully to

compress air to a pressure which was higher than the intended value of 250 bar. The

discharge pressure managed to be raised up to 260 bar. During the tests, each of part

of the compressor was working properly and showed the satisfactory in relation to

the specified requirement. Figure 5.24 to Figure 5.27 have show the performance of

97

the compressor (or piston ring assembly) as the pressure of after each stage is raised

gradually until steady state condition was arrived.

Figure 5.24 and Figure 5.25 show that the compressor has produced the

pressure around 230-240 bar. The maximum pressure obtained for first test was only

240 bar at the end of test, but the time taken was about 900 second which was too

long. The compressor was stopped to investigate condition of all mechanical parts

and to look at any damaging effect of testing inside the compressor casing. When

every moving parts was found to be in good condition, the second test was carried

out and produced the high pressure of 240 bar at a relatively longer 1700 second to

reach steady state. This result can be seen in Figures 5.25. More investigation was

needed to find this problem of long transient time. After this second test, it was found

that there was a technical problem at the pressure relief valve which was not properly

calibrated. Although set at 250 bar the relief valve opened at 240 bar.

After the setting of pressure relief valve problem was overcome, a fourth test

was conducted. The results are given in Figure 5.26. During this test, the maximum

pressure was surprisingly higher and reached a steady value of around 260 bar. This

test was continued with a longer time and the piston ring assembly seemed to be able

to function at steady state at 260 bar of final discharge pressure, as shown Figure

5.27. Through these testings, it could be concluded that the main objective of the

study was finally achieved, which is to prove that the material selection, design,

analysis of piston ring assembly is able to compress air up to 250 bar, while at the

same time the set specifications of the new symmetrical wobble-plate compressor are

met.

CHAPTER VI

CONCLUSIONS AND RECOMMENDATIONS

6.1 Conclusions

In this research, Piston Ring Assembly (PRA) for a new symmetrical wobble-

plate compressor for Home Refueling Appliance (HRA) was designed, developed,

and successfully tested. The significant outcomes of this study can be summarised as

follows:

1. Piston Ring Assembly (PRA) is a critical component which contributes to the

majority of frictional losses in a reciprocating compressor where the selection

of material for Piston Ring Assembly (PRA) is very important to meet the

specified compressor performance.

2. PTFE and PEEK materials of piston rings are suitable for this oil-free

lubrication for a new symmetrical wobble-plate compressor.

3. Coefficients of friction of PTFE and PEEK sliding against XW41 hard

chromed have values ranging from µ = 0.20 to µ = 0.22 and µ = 0.29 to µ =

0.37 respectively.

4. Wear rate of PTFE was found to be between 0.02µm/hr to 0.12µm/hr and that

of PEEK was between 0.02µm/hr to 0.10µm/hr.

99

5. The new symmetrical wobble-plate reciprocating compressor prototype and

its piston ring design, material and assembly have been tested successfully to

compress air to extremely high pressure of 260 bar.

6.2 Recommendations for Future Research

This research has carried out work on the development of Piston Ring

Assembly (PRA) for a new symmetrical wobble plate compressor. However there are

still several aspects could be done or continued to further develop the new

compressor in term of performance:

1. Perform endurance test on the Piston Ring Assembly (PRA) to predict the

operating life.

2. In order to improve performance of the piston ring, test on the polymer as a

self lubricated material should be conducted with other fillers combinations.

3. The tribotest on friction and wear should be tested under various tribological

environments that practically influence the performance of Piston Ring

Assembly (PRA) e.g. speed, temperature, surface roughness, time or distance.

REFERENCES

Affenzeller, J. and Gläser, H. (1996). Bearings and Lubrication of Internal

Combustion Engines In German. Vienna, Austria, Springer-Verlag. 397 p.

ISBN 3-211-82577-0.

Ahmad, Z. A. (2006). Structural Design and Analysis of A New Symmetrical

Multi-Stage Wobble-Plate Compressor. MEng, Thesis Universiti Teknologi

Malaysia.

American Petroleum Institute (2005). API 618: Reciprocating Compressors for

Petroleum, Chemical, and Gas Industry Services. America: American

Petroleum Institute.

Baumann, H. (1994). Design Features of a Small Oil-Free, Reciprocating, High

Pressure Compressor. Proceedings International Compressor Conference

Purdue University: 381-385.

Baumann, H. and Conzett, M. (2002). Small Oil Free Piston Type Compressor for

CO2. Proceedings International Compressor Conference Purdue University.

Blau, P. J. (1999). Fifty Years of Research on the Wear of Metals. Elsevier

Science, Tribology International Volume 30 Number 5.

Bloch, H and Hoefner, J. (1996). Reciprocating Compressors:

Operation and Maintenance. Butterworth-Heinemann.

101

Bottomley, J. W. (1994). Application of New Sealing Component Materials To

Reciprocating Compressors. IMechE Journal, C477/013/1994.

Chui, B, K. (2001). Computational Analysis of Piston Ring Wear and Oil

Consumption for an Internal Combustion Engine. MsC Thesis. Michigan

State University,

Chlumsky, V. (1965). Reciprocating and rotary compressor. SLTN Publisher of

Technical Literature. Prague, Czechoslovakia.

Dunaevsky, V. (2001). The Effect of Contact Pressure on Piston Ring Twist.

SAE Paper No. 2001-01-2720.

Dunaevsky, V. (1999). Analysis of Elastic Distorsions of a Piston Ring in the

Reciprocating Air Brake Compressor Due to the Installation Stresses. SAE

Paper No. 1999-01-3770.

Dwivedi, S. N. (1990). Design Consideration for High Pressure Reciprocating

Compressor for Refinery services. International Compressor Engineering

Conference at Purdue pp 273-281.

Gupta, P. (2001). On a Multi-Process Wear Model. Lubrication Engineering.

57(2001)4, pp. 19–24.

Hanlon, P.C. (2001).Compressor Handbook. New York: McGraw-Hill.

Haubner, F. G. (2001). Simulation of the Piston Ring Dynamis. SAE Paper No.

2001-01-3368.

Hariu, N., Nakayama, K., Hill, W.R. (1991). Dynamic Analysis Of A Variable

Capacity Compressor For A Vehicle Air-conditioning Using Finite Element

Methods. SAE Paper, No. 910213.

102

Ishi, N., Abe, Y., Taguchi, T., Maruyama, T. and Kitamura, T. (1990). Dynamic

Behaviour of Variable capacity Wobble-plate Compressor for Automotive Air

Conditioners. Proc. Int. Compressor Conf., Purdue University: 345-353.

Kauzlarich, J. J. and Williams, J. A. (2001). Archard Wear and Component

Geometry. Proceedings of the Institution of Mechanical Engineers Part J.

Journal of Engineering Tribology, 215-4, pp. 387–403.

Maczek, K. and Wolek, W. (1994). Confinement and Advoidance of Lubricants in

Reciprocating Compressors. Proceedings International Compressor

Conference Purdue University.

Mid-Michigan Research. (1998). CASE Theoretical Manual. Mid-Michigan

Research.

Musa, M.N. (2005). Wobble Plate Compressor. (PI 2005 5456).

Nishikawa, H. and Nishikawa, T. (1998). Development of the Non-Lubricated Four-

Stage Compressor Compressing Up To 24.25MPa. Proceedings International

Compressor Conference Purdue University: 183-188.

Nishikawa, T., Nishikawa, H., Obokata, T. and Ishima, T. (2000). A Study for

Improvement on High Pressure Multi-stage Reciprocating Compressor.

Proceedings International Compressor Conference Purdue University: 105-

112.

Ouwerkerk, C. and Theeuwes, G. (1981). Leakage and Friction of Different Dry-

Running Piston Ring. 9th International Conference on Fluid Sealing. Paper

F1.

Poole, E. J. (1978). Development Since 1950 of Standard Range Industrial Air

Compressor. IMechE C16/78.

103

Priest, M. and Taylor, C. M. (2000). Automobile Engine Tribology - Approaching the

Surface. Wear, 241-2, pp. 193–203.

Radcliffe, C. (2005). Sealing Material Developments for Reciprocating Gas

Compressor. Sealing Technology Journal.

Riegger, O.K. (1990). Design Criteria and Performance of an Advance

Reciprocating Compressor. Proceedings International Compressor

Conference Purdue University: 61-68.

Sherrington, I. and Mercer, S. (2000). The Use of Topography-Based Parameters for

The Assessment and Prediction of Surface Wear. Tribotest 7, pp. 3–11.

Shivakant and Krishna. (2001). Mechanics and Stress Analysis of Piston Ring In

Multibody Single Cylinder Internal Combustion Engine-FE Analysis.

Proceedings International Compressor Conference Purdue University.

Shuster, M., Mahler, F., and Crysler, D. (1999). Metallurgical and Metrological

Examinations of The Cylinder Liner - Piston Ring Surfaces After Heavy Duty

Diesel Engine Testing. STLE Tribology Transactions, 42, pp. 116–125.

Smith, E. H. (1994). Mechanical Engineer’s Reference Books. Twelfth Edition

Butterworth-Heinemann.

Society of Automotive Engineers (SAE). (1969). Metallic Seal Rings for High

Temperature Reciprocating Hydraulic Service. SAE-AIR 1077.

Syahrom, A. (2006). Design and Development of Multi-Stage

Symmetrical Wobble-Plate Compressor. MEng. Thesis. Universiti Teknologi

Malaysia.

Tian, C.Q., Dou, C.P., Yang, X.J., Li, X.T. (2004). A mathematical model of

Variable displacement wobble-plate compressor for automotive air

conditioning system. Applied Thermal Engineering Vol. 24: 2467-2486.

104

Todsen, U and Kruse, H. (1982). Hydrodynamic Optimisation of The Tribological

System Piston-Ring-Liner. Proceedings International Compressor Conference

Purdue University.

Tojo, K., Takao, K., Ito, M. and Hayase, I. (1990). Dynamic Behaviour of variable

capacity compressor for automotive air conditioners. SAE Paper, No.

900084.

Wilson, R. S. (1990). Development in Piston and Rod Sealing Material for Oil-free

Gas Compressor. IMech E. Fluid Machinery Congress The Hague.

Wrede, F. (1978). Comparison of Mathematical and Experimental Simulation of

Lubrication Condition In the System of Piston, Piston Ring and Liner.

Proceedings International Compressor Conference Purdue University.

Yong, L. (1986). Prediction for The Sealing Characteristics of Piston Rings of a

Reciprocating Compressor. Proceedings International Compressor

Conference Purdue University.

105

APPENDIX A

Piston Groove Dimension of Piston Ring

(According to ISO 7425/1)

106

Figure A.1 Nomenclatures of piston groove dimension of piston ring

Table A.1 ISO 7452/1 Standard for piston groove dimension of piston ring

Table A.2 Result piston groove dimension of piston ring for Home Refueling Appliance (HRA) compressor

Piston Bore

Diameter Groove

Diameter Groove Width Radius

Radial Clearance

Piston Diameter

O-Ring Dimension

Lead-in Chamfer

DN H9 d1 h9 L1 +0.2 r1 Smax d2

HRA_1 36 28.5 3.2 0.6 0.4 35.6 28.24 x 2.62 1.4 HRA_2 24 16.5 3.2 0.6 0.4 23.6 15.54 x 2.62 1.4 HRA_3 15 7.5 3.2 0.6 0.4 14.6 7.59 x 2.62 1.4 HRA_4 10 5.1 2.2 0.4 0.2 9.8 4.8 x 1.8 1.1

107

APPENDIX B

Piston Groove Dimension of Rider/ Guide Ring

(According to ISO 7425/1)

108

Figure B.1 Nomenclatures of piston groove dimension of rider/ guide ring

Table B.1 ISO 7452/1 Standard for piston groove dimension of rider/guide ring

Table B.2 Result piston groove dimension of rider/ guide ring for Home Refueling Appliance (HRA) compressor

Piston Bore

Diameter Groove

Diameter Groove Width Thickness Radius

Radial Clearance

Ring Gap

DN H9 d2 h8 L2 +0.2 W rmax S1 Z

HRA_1 36 31 5.6 2.5 0.2 0.4 2

HRA_2 24 19 5.6 2.5 0.2 0.4 2

HRA_3 15 11.9 4 1.55 0.2 0.4 2

HRA_4 10 6.9 4 1.55 0.2 0.2 2

109

APPENDIX C Material properties of Piston Ring Assembly (PRA):

110

Table C.1 Material properties of piston and cylinder liner:

Material

Density (gr/cc)

PoissonRatio

Young’s Modulus

(GPa)

Yield Strenght (MPa)

Hardness Vickers

Melting Point (°C)

AISI 304 8 0.29 193 215 129 1400-1455

Characteristics:

Austenitic Cr-Ni stainless steel. Better corrosion resistance than Type 302. High

ductility, excellent drawing, forming, and spinning properties. Essentially non-

magnetic, becomes slightly magnetic when cold worked. Low carbon content means

less carbide precipitation in the heat-affected zone during welding and a lower

susceptibility to inter-granular corrosion.

Applications:

Chemical equipment, coal hopper linings, cooking equipment, cooling coils,

cryogenic vessels, dairy equipment, evaporators, flatware utensils, feed water tubing,

flexible metal hose, food processing equipment, hospital surgical equipment,

hypodermic needles, kitchen sinks, marine equipment and fasteners, nuclear vessels,

oil well filter screens, refrigeration equipment, paper industry, pots and pans,

pressure vessels, sanitary fittings, valves, shipping drums, spinning, still tubes, textile

dyeing equipment, tubing.

111

Table C.2 Material properties of piston ring (PTFE):

Material

Density (gr/cc)

PoissonRatio

Young’s Modulus

(GPa)

Yield Strenght(MPa)

Hardness Vickers

Melting Point (°C)

PTFE

40% Bz

3.13 0.33 0.46 28 67 260

Characteristics:

This material exhibits a unique combination of heat resistance and low friction

together with outstanding chemical and good electrical properties. Continuous use

temperatures range -204°C to +260°C, no moisture absorption, high arc resistance,

and is self lubricating with a low coefficient of friction. Compounds of bronze and

Teflon improve creep strength, thermal conductivity and wear resistance of moldings

as well as low friction.

Applications:

Useful in applications which undergo high mechanical loads or high-speed rubbing

contacts where the bronze filler supplies the strength and conductivity to carry away

excess, unwanted heat, e.g. piston ring, rider ring, journal bearing, valve plate, etc.

112

Table C.3 Material properties of piston ring (PEEK):

Material

Density (gr/cc)

PoissonRatio

Young’s Modulus

(GPa)

Yield Strenght(MPa)

Hardness Vickers

Melting Point (°C)

PEEK 1.44 0.4 8.3 110 130 344

Characteristics:

PEEK exhibits excellent resistance to a wide range of organic and inorganic

chemicals. Polyethetetherketone ( PEEK ) is a high performance thermoplastic with

the characteristics common to this group - strong, stiff, hard, high temperature

resistance, good chemical resistance and inherently low flammability and smoke

emission. PEEK is pale amber in color and usually semi-crystalline and opaque,

except thin films are usually amorphous and transparent. It also has very good

resistance to wear, dynamic fatigue and radiation, but it is difficult to process and

very expensive.

Applications:

Applications include flexible printed circuit boards (film), piston ring, rider ring,

journal bearing, valve plate, monofilaments, injection molding engineering

components and items used in aerospace and radiation environments. etc.

113

APPENDIX D

Patent Filing No: PI 2005 5056

(Wobble Plate Compressor)

114

115

116

117

WOBBLE PLATE COMPRESSOR

The present invention relates to wobble plate compressor design. More particularly,

the present invention relates to an improved wobble plate type compressor or swash

plate compressor having a symmetrical and multistage configuration for use in a gas

compression system.

BACKGROUND TO THE INVENTION

Gas compression systems required to increase the gas pressure are well known.

Gas pressure need to be increased for gas transmission purposes and for storage

purposes. Gas can only be distributed when pressure difference exist. For gas

storage, gas pressure need to be increased to reduce the amount of volume

required to store the gas. Pressure required is high to be achieved by just using

single stage compression thus multistage compression need to be used.

Reciprocating piston compressor is a natural choice for high pressure and small to

medium flow rate requirement. Piston compressor has many variances according to

the piston arrangement and chosen driver mechanism. Crankshaft drive compressor

can be found with inline, V-shape, L-shape, vertical, horizontal and radial piston

arrangement. Coaxial piston arrangement can also be achieved using swash plate

and wobble plate mechanism.

Wobble plate compressor has long been used in the automotive air conditioning

system with single stage compression. Example of fixed capacity wobble plate

compressor is given in U.S. Patent No. 4784045 while variable capacity wobble

plate compressor is given in U.S. Patent No. 4428718. Variable capacity wobble

plate compressor has the ability to change its capacity by changing the piston stroke

through varying the wobble plate tilting angle. The ratio between discharge and

crankcase pressure is used to control wobble plate tilting angle. Connecting rod is

used to connect wobble plate with piston with ball joint interface at both ends. Some

of the inventions as in U.S. Patent No.5079996 omitted ball joint connection at

piston side due to small piston depth or small wobble plate tilting angle as in U.S.

Patent No. 4138203. Wobble plates slide on rotor either by using roller bearing and

118

thrust bearing, thrust bearing and spherical bearing or roller bearing only as found in

U.S. Patent No. 4867649, 4869651 and 4138203 respectively. Wobble plate is

prevented from rotating with rotor using anti rotation mechanism which is either

thrust rider and slider plate (U.S. Patent No. 3552886), ball and slider plate (U.S.

Patent No. 4105370), ball and guide rod mechanism (U.S. Patent No. 5094590),

Rzeppa mechanism (U.S. Patent No. 5079996) and bevel gear (U.S. Patent No.

4869651). Slanted or fully supported drive shaft at both ends has been used in all

previous invention. Rotor shapes for fixed capacity wobble plate compressor tend to

be simpler as found in U.S. Patent No. 4869651 whereby for variable capacity

wobble plate compressor, some arrangement is needed to change wobble plate

tilting angle with the typical design is as found in U.S. Patent No. 4428718. Many

improvements or design variation has been made on this mechanism alone.

Housing design is normally split into two and three piece parts with cylinder block

imbedded into the housing. End plate is used to house valve plate and lubrication

pump.

SUMMARY OF THE INVENTION

Accordingly, it is an object of the current invention to provide a wobble plate type

compressor aligned in symmetrical configuration, which results in significant

reduction of vibration of the compressor. The pistons, cylinder block and wobble

plate which mirror each other at the centre of drive shaft reduces if not eliminates

the horizontal force caused by gas reaction forces acting on the pistons.

Another object of the current invention is the introduction of multistage configuration

that will allow higher gas compression than normally attained in single stage wobble

plate compressors used in automotive refrigeration.

A further object of the current invention is to provide a wobble plate of the character

described which is particularly designed to compress gas without oil contamination.

This feature eliminates the inevitable blow-by of oil vapor passing into the gas being

compressed, as the present feature is free from lubricating requirements on the part

of the operator between periodic maintenance.

Yet another object of the current invention is to provide a compressor of the

characters above which will involve a fewer number of parts with reduced machining

119

requirements, and which may be easily and rapidly assembled to provide a unit at

minimum cost.

Yet another object of the present invention is to provide a wobble plate compressor

of the characters described which is composed of durable parts affording easy

disassembly when required for maintenance and affording long, useful life.

Still another object of the present invention is to provide a structure of the character

described which may be scaled up or down to readily provide units of different sizes

and capacities and also be adopted to swash plate type compressors.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will now be described in greater detail, by way of an example, with

reference to the accompanying drawings, in which:

Figure 1 is the isometric view of the compressor illustrating the housing, left and

right end plate and acrylic cover;

Figure 2 is a see through drawing illustrating the assemblies of the internal

component in the compressor;

Figure 3 is a section view showing parts of wobble plate and anti-rotation

mechanism;

Figure 4 is the isometric view of piston assemblies illustrating different size and

shapes of pistons and couplers;

Figure 5 is a cross section view of wobble plate assemblies illustrating the

components involved;

Figure 6 illustrates the arrangement of cylinder block at the end plate, shown here

with possible fittings arrangement at the suction and discharge port at each cylinder

block;

Figure 7 is a cross section view of piston assemblies illustrating the components

involved;

120

Figure 8 is a cross section view of cylinder block illustrating arrangement of liner,

valve cover, suction valve plate and discharge valve plate; and

Figure 9 is cross section view of cylinder block illustrating the suction port and

discharge port.

DETAILED DESCRIPTION OF THE DRAWINGS

Referring to Figure 1 through Figure 9, wherein like numerals indicate like

corresponding parts throughout the nine views, a symmetrical multistage wobble

plate compressor is generally shown.

The compressor has a housing, which includes crank case 1, left end plate 2, right

end plate 3 and acrylic cover 4. Left end plate 2 and right end plate 3 are clamped to

crank case 1 using bolts.

The compressor of the present invention comprises of two sets of pistons in cylinder

blocks 5, wobble plate 6, rotor 7 that mirror each other as shown in Figure 2. Drive

shaft 8 is stepped at both ends to locate and fix the bearing at the end plate 2, 3 at

both ends. Drive shaft 8 and rotor 7 are fixed together using pin. Rotor 7 and wobble

plate 6 is connected together through deep groove bearing 9. Wobble plate 6 have

slots for wobble plate pins 10, slot for anti-rotation ball 11 and flange at the front

face around the periphery of bearing slot at its centre. Rotor 7 is provided with a slot

for pin and flange at the back face.

Bearing 9 is tight fitted to both rotor 7 and wobble plate 6. External c-clip is used to

secure rotor 7 with bearing 9 while internal c-clip is used to secure wobble plate 6 with bearing 9. This will prevent wobble plate 6 or bearing 9 from sliding to the front

in case of tight fit failure. Flange at the front face of wobble plate 6 will press against

the front face of bearing outer race 12 whereas flange at the back of rotor 7 will

press against the back face of bearing inner race 13. This will ensure bearing 9 or

wobble plate 6 from sliding to the back in case of tight fit failure.

121

Rotation of drive shaft 8 with rotor 7 will induce wobbling motion in the wobble plate

6 through the bearing 9 interface. Wobble plate 6 is prevented from rotating with

rotor 7 by the anti rotation mechanism which consist of a guide rod 14 and hollow

spherical ball 11 that slide horizontally on the guide rod 14 and up and down in the

slot for anti-rotation ball at wobble plate 6. Wobble plate 6 wobbling motion will be

transferred into piston 15 reciprocating motion through connecting rod 16. End joint

connection is used as the interface between the connecting rod 16 and wobble plate

6 and between the connecting rod 16 and piston 15.

Different piston diameter size 15, 17, 18, 19, 20 and its corresponding cylinder block

5, 26, 27, 28, 29 are used for each stage. The largest piston 15 and cylinder block

diameter size 5 is for the first stage. Piston and cylinder block diameter size will

correspondingly reduce for higher number of stage. Each piston set has different

shape of pistons 15, 17, 18, 19, 20 with corresponding number of groove, piston

rings/rider rings and coupler 21, 22, 23, 24, 25. The variations depend on stage

pressure involved.

Pistons for the first stage to the third stage 15, 17, 18 are made from aluminum

while the fourth and fifth stage 19, 20 is made from hard steel. Liner 39 is made from

cast iron. The inner surface of the liner is hard-chromed to obtain mirror surface

finishes. Piston ring and rider ring is made from self-lubricated PTFE material.

Labyrinth groove is used for the forth stage and fifth stage piston omitting piston

rings due to small piston diameter size. Teflon material is used for the liner at the

last two stages with the clearance between piston and cylinder block is 5µm.

Coupler 21 is used to connect piston 15 to their respective connecting rod 16 using

end-joint 30, 31. Holes are made at the coupler 21 to ensure mass of each piston 15

with its corresponding coupler 21 is the same for all stages. Bolt 32 is used to fixed

pistons 15 with coupler 21. Couplers 21 are fitted to connecting rod end-joint ball 33

at the piston side using piston pin 34, which is secured in the coupler 21 using two

internal c-clips. End-joint ball centre is located on pin using piston bush 35.

Connecting rod 16 is composed of female 31 and male end-joint 30 that are

screwed into each other. A connecting rod 16 length is determined by length of

thread engagement between both end-joint and fixed using nut and thread lock.

Connecting rod end-joint ball at the wobble plate 36 is also fitted to wobble plate 6

122

using wobble plate pin 37, which is secured in the wobble plate 6 using internal c-

clip. End-joint ball centre is located on wobble plate pin 11 using wobble plate bush

38.

All the bearing and end-joint used is lubricated using grease that needs no

maintenance within the periodic maintenance interval. Sealing between cylinder

block 5, liner 39 and valve seat 40 is achieved using o-ring. O-ring is also used

under piston ring to press piston ring against liner bore surface.

Each cylinder block 5 has two ports for suction 41 and discharge 42. Suction valve

plate 43 is positioned between liner 39 and valve cover 44 while discharge plate

valve 45 is positioned between valve cover 44 and valve seat 40. Fins on cylinder

block 5 are used for cooling purpose.

123

CLAIMS 1. In wobble plate type compressor comprising of a compressor housing having

a cylinder block (30) provided with a plurality of cylinders 5, 26, 27, 28, 29 and a

crank chamber 46 enclosed within each of said cylinders which is free from

contamination, a drive shaft 8 rotatably supported in said housing, a rotor 7 around

the drive shaft 8 which is arranged back to back and further connected to an inclined

wobble plate 6, a coupling member of said wobble plate with each having a plurality

of pistons arranged symmetrically, said coupling member having one end which is

coupled with said wobble plate and another end which is coupled with each of said

symmetrical pistons, and a rotation preventing means for preventing rotation of said

wobble plate with the rotor.

2. The wobble plate compressor as claimed in claim 1, wherein said

compressor is multistage in design with symmetrical back-to-back arrangement that

reduces vibration and sound.

3. The wobble plate compressor as claimed in claim 2, wherein said multistage

design provides for different pistons and cylinder block sizes at different stages

allowing for higher gas pressure compression.

4. The wobble plate compressor as claimed in claim 1, wherein said wobble

plate coupling member comprises of a connecting rod with end-joint connection at

both ends to connect between pistons and wobble plate.

5. The wobble plate compressor as claimed in claim 1, wherein said piston

rings and rider rings are self lubricated, while said bearings and end joints are

equipped with grease for lubrication, thus providing a contamination free wobble

plate.

6. The wobble plate compressor as claimed in claim 1, wherein said rotation

preventing means prevents rotation of wobble plate with rotor by the anti rotation

mechanism which consist of a guide rod and hollow spherical ball that slide

horizontally on the guide rod and up and down in the slot within the wobble plate.

124

ABSTRACT

WOBBLE PLATE COMPRESSOR

A wobble plate type compressor for vehicles having a symmetrical and multistage

configuration is disclosed which includes two sets of plurality different pistons size

being reciprocated within respective cylinders by two wobble plate members that

mirror each other. Multistage configuration will have multiple piston diameter size

that allow for higher gas pressure compression.

4 3

1

2

Figure 1

125

Figure 2

Figure 3

11

14

6

7 10

9

5

8

126

16

Figure 4

Figure 5

15

17

18

19

20

6

7

9 21

25

22 24

23

12 6

7 9

13

127

5 2

26

29

27 28

Figure 6

16

Figure 7

15

21 34

35 37

38

30

36

32 33 31

128

40 5

43

Figure 8

Figure 9

46

45

44

42

41