. 00‘ wfl'fiw<t-. ..-. --< .< .-. .«~< . .,. .q. .

D .[ ,lul'il’.|rl|b

.r "'|

11»:

ICHI IGANS STATE

I IIII lIIIIIIIIIIII.’ IIIIIIIII

1293 00882 5741

III

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled

DEVELOPMENT AND ANALYSIS OF A FIRE
RING WEAR MODEL FOR A PISTON ENGINE

presented by

Yooseok Chung

has been accepted towards fulfillment
of the requirements for

Ph.D. Mechanical Engineering

degree in

 

 

 

 

é/W y. JM/

7
Major professor

 

Date 8/ 3 /92

 

MSU is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

 

 

LIBRARY
Michigan cram
' University

P..-

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

   
  

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

fl
_JE::IL_I
flIf?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DEVELOPMENT AND ANALYSIS OF A FIRE RING WEAR
MODEL FOR A PISTON ENGINE

By

Yooseok Chung

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Mechanical Engineering

1992

ABSTRACT

DEVELOPMENT AND ANALYSIS OF A FIRE RING WEAR MODEL
FOR A PISTON ENGINE

By

Yooseok Chung

Ring wear may not be a problem in most current automotive engines, however
a small alteration in the ring face geometry can significantly affect the hydrodynamic
lubrication characteristics of the ring. This in turn can cause excessive frictional loss-
es and blowby in an engine. As engines become more compact and highly loaded, ring
wear is likely to be more severe than in current engines. To be able to assess the ef-
fect of ring loading, a piston ring wear model has been developed through the use of
ring dynamics analysis with the assumption of a linear relationship between ring wear
and the friction work applied on the surface of the ring. It is also assumed that ring
wear occurs only in the mixed or boundary lubrication regime, where the hydrodynam-
ic film breaks down. The lubrication regimes are separated by the nominal minimum
film thickness, which is defined by the ratio of the minimum oil film thickness to the ef-
fective surface roughness of the joint.

Input dynamics to the piston ring wear model were provided by a commercially
available ring dynamics analysis program which has been under development for the
past five years. Analysis shows that the minimum oil film thickness at a given engine
speed tends to be lower just after 'I'DC and BDC, which implies that mixed lubrication
occurs between the piston ring and the cylinder wall at these points. This effect ap-

pears more prominent at the lower engine speeds than at the higher engine speeds

under the same power output. The equation developed for the piston ring wear model
is based on the several results of experimental wear efforts. This ring wear analysis
clearly shows that the higher the engine speed, the lower the wear rates at the same
power output, as is obtained through the experiment. The Specific wear rate at steady
state drawn from this analysis corresponds to the maximum value of the wear coeffi-
cient observed in other experimental studies

The present analysis was developed and the predictions were compared to da-
ta available for a Diesel engine. The model developed in this study can be applied to
any piston engine, as long as the Specifications of the engine such as the oil viscosity,
temperature distributions in the cylinder bore, the ring geometry, the ratio of cylinder
bore hardness to ring hardness, and the ring face profile are given as the same. This
is the first model of piston ring wear which includes ring dynamics, and is intended to
provide a framework for a general model which can include the factors mentioned

above. This ring wear model can be used to estimate apriorr' wear rates for R11 or

SLA studies.

To My Parents

iv

ACKNOWLEDGMENTS

I would like to express my appreciation to my adviser, Dr. Harold Schock, for
his inspiring guidance and encouragement during this dissertation work. His priceless
suggestions and insights have greatly enriched this research.

Dr. Craig Somerton is due special thanks for putting forth a special effort in lis-
tening to a number of my presentations and providing helpful comments.

I am also grateful to my other committee members, Dr. David Grummon and
Dr. William McHarris, for their helpful suggestions.

Other who have provided valuable comments throughout this dissertation
work include Dr. L. Brombolich, Dr. K. Lee, Dr. R. Ronningen, Mr. S. Yoo, and Mr. R.
Schafer.

Finally, I thank my parents and my lovely wife for their love and continuous

support throughout my graduate studies.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................ ix
LIST OF FIGURES ................................................................................................ x
CHAPTER 1 - INTRODUCTION ..................................................................... 1
l. 1 Problem Statement .............................................................. 1
1. 2 Dissertation Arrangement .................................................... 2
CHAPTER 2 - BACKGROUND ......................................................................... 4
2. 1 Piston Ring Design ................................................................ 4
2. 2 Piston and Ring Assembly Friction ....................................... 7
2. 3 Lubrication and Oil Consumption ......................................... 9

2. 4 Types of Wear ....................................................................... 11

2. 5 Experimental Studies on Ring Wear Measurement ............... 12

2. 5. 1 Ring Wear Measurement using SLA ......................... 12

2. 5. 2 The Effect of Ion Implantation on Friction and Wear 12

2. 5. 3 Radioactive Ion Implantation Technique .................. 13

CHAPTER 3 - GAS FLOW ANALYSIS ............................................................. 14
3. 1 Introduction ............................................................................. 14

3. 2 The Mass Flow Rate ................................................................ 18

3. 3 The Determination of Inter-Ring Gas Pressure ...................... 26

3. 4 The Axial Motion of the Fire Ring in the Groove ................... 27

vi

3. 5 Results and Discussion ......................................................... 29

CHAPTER 4 - THE ANALYSIS OF FIRE RING FRICTION ........................ 40
4. l Piston Kinematics .................................................................. 40

4. 2 The Indicated Mean Effective Pressure and Brake Power ..... 43

4. 3 Reynolds Equation ................................................................ 45

4. 4 Lubrication ............................................................................ 50

4. 5 Fire Ring Friction ................................................................. 51

4. 5. 1 Hydrodynamic Lubrication ....................................... 53

4. 5. 2 Mixed and Boundary Lubrication .............................. 54

4. 6 Results and Discussion ........................................................... 55

CHAPTER 5 - THE APPROXIMATE MODEL OF FIRE RING WEAR ...... 65
5. 1 Relationship between Friction and Wear .............................. 65

5. 2 Fire Ring Wear Model .......................................................... 66

5. 3 Results and Discussion .......................................................... 72

5. 3. 1 Other Work related to Wear Studies .......................... 72

CHAPTER 6 -

CHAPTER 7 -

5. 3. 2 Error Estimation and Discussion of

Predicted Results ........................................................ 74
SUMMARY AND CONCLUSIONS .......................................... 83
RECOMMENDATIONS ........................................................... 88
7. 1 Limitations of the Model ....................................................... 88
7. 2 Recommendations for Future Research ................................ 88

vii

APPENDIX A - THE APPLICATION OF RADIOACTIVE ION IMPLANT A-
TION TECHNIQUES FOR PISTON RING WEAR MEA-

SUREMENT .......................................................................... 90

A. 1 The Advantages of R11 over SLA ........................................ 90

A. 2 The Measurement Procedure of Piston Ring Wear .............. 91

A. 3 Radioactivity ......................................................................... 93

A. 4 The Sensitivity Analysis of Measurement ........................... 96

APPENDIX B - THE TRIM SIMULATION ...................................................... 100
B. 1 Introduction ........................................................................... 100

B. 2 Physical Assumptions used in TRIM .................................. 101

B. 3 Preliminary Simulations and Results .................................... 102
BIBLIOGRAPHY ..................................................................................................... 106

viii

LIST OF TABLES

Table 1 Specifications of the engine ......................................................................... 31
Table 2 Constants in the Vogel Viscosity Equation .................................................. 51
Table 3 IMEP and power output under different engine operating conditions .......... 55
Table 4 Wear coefficients .......................................................................................... 77

Figure 2.1
Figure 3.1
Figure 3.2
Figure 3.3
Figure 3.4
Figure 3.5
Figure 3.6
Figure 3.7
Figure 3.8

Figure 3.9

Figure 3.10
Figure 3.11

Figure 3.12

Figure 4.1
Figure 4.2

Figure 4.3
Figure 4.4

Figure 4.5
Figure 4.6
Figure 4.7
Figure 4.8

LIST OF FIGURES

Schematic of piston and ring assembly in a piston engine ................... 5
Detailed schematic of clearance volume in a piston engine ................. 15
Orifice volume model of ring pack ...................................................... 17
Schematic illustration of energy balance .............................................. 18
Procedure for the determination of inter-ring gas pressures ................. 35
Gas pressures in each region at an engine speed of 1200 rpm ............ 36
Gas pressures in each region at an engine speed of 1500 rpm ............ 36
Gas pressures in each region at an engine speed of 1800 rpm ............ 37
Gas pressures in each region at an engine speed of 2100 rpm ............ 37
Ring position in the groove at an engine speed of 1200 rpm ............ 38
Ring position in the groove at an engine speed of 1500 rpm ............ 38
Ring position in the groove at an engine speed of 1800 rpm ............ 39
Ring position in the groove at an engine speed of 2100 rpm ............ 39
The geometry of a reciprocating piston engine ..................................... 41
Geometry of lubricated junction between a piston ring

and cylinder bore .................................................................................. 49
Stribeck diagram for a piston ring ........................................................ 52
Piston location in a diesel engine .......................................................... 56
Piston velocity at each engine speed .................................................... 56
Piston acceleration at each engine speed ............................................... 57
P-V diagram at an engine speed of 1200 rpm ....................................... 58
P-V diagram at an engine speed of 1500 rpm ....................................... 58

Figure 4.9 P—V diagram at an engine speed of 1800 rpm ..................................... 59

Figure 4.10 P—V diagram at an engine speed of 2100 rpm ..................................... 59
Figure 4.11 Minimum oil film thickness at an engine speed of 1200 rpm ............. 61
Figure 4.12 Minimum oil film thickness at an engine speed of 1500 rpm ............. 61
Figure 4.13 Minimum oil film thickness at an engine speed of 1800 rpm ............. 62
Figure 4.14 Minimum oil film thickness at an engine speed of 2100 rpm ............. 62
Figure 4.15 Ring friction force at an engine speed of 1200 rpm ............................ 63
Figure 4.16 Ring friction force at an engine speed of 1500 rpm ............................ 63
Figure 4.17 Ring friction force at an engine speed of 1800 rpm ............................ 64
Figure 4.18 Ring friction force at an engine speed of 2100 rpm ............................ 64
Figure 5.1 Schematic illustration of true area of contact ...................................... 68
Figure 5.2 Ring wear rate for one cycle of engine operation at 1200 rpm ........... 79
Figure 5.3 Ring wear rate for one cycle of engine operation at 1500 rpm ........... 79
Figure 5.4 Ring wear rate for one cycle of engine operation at 1800 rpm ........... 80
Figure 5.5 Ring wear rate for one cycle of engine operation at 2100 rpm ........... 80
Figure 5.6 Accumulated ring wear for one cycle of engine operation ................... 81
Figure 5.7 Theoretical average ring wear rate at each engine speed .................... 82
Figure 5.8 Experimental average ring wear rate at each engine speed ................. 82
Figure 6.1 Procedure for fire ring wear analysis ................................................... 86
Figure A.1 Activation of the piston ring by radioactive ion beam ......................... 92
Figure A.2 Idealized uniform dose-depth distribution .......................................... 92
Figure A.3 Detector efficiency calibration ............................................................ 95
Figure A.4 Normal distribution ............................................................................. 97
Figure B.1 The ion range profile of a 2.5 MeV/amu 20Ne beam implanted
onto Si3N4 ............................................................................................. 104

Figure B2 The radiation damage of Si3N4 for a 2.5 MeV/amu 20Ne beam ......... 104

xi

Figure 3.3 Polyenergetic profile for a 2.5 MeV/amu 20Ne beam degraded by
stacked gold foils .................................................................................. 105

xii

CHAPTER 1

INTRODUCTION

1. 1 Problem Statement

Piston rings in a piston engine play important roles in the performance and en-
durance of the engine, viz., sealing the combustion gas between the piston and cylin-
der wall to prevent excessive blowby, controlling the amount of lubricating oil deposit-
ed on the cylinder wall as well as working as the passage of heat flow from the piston
into the cylinder. Yet it must perform this task with minimum frictional resistance to
avoid excessive energy loss and with minimum wear to ensure an acceptable life of
the piston and cylinder liner. These requirements must be satisfied under severe con-
ditions of high temperature and pressures and cyclic variations of loads, speeds, and
lubricant viscosity.

Wear within design limits (normal wear) is acceptable from the viewpoint of
both safety and economics, since the frictional compatibility of the ring is often im-
proved during an initial stage of the engine operation. However, after a substantial pe-
riod of running-in, the mean wear rate experienced by a piston ring should be suffi-
ciently low to provide a satisfactory life for the engine. This long term wear rate would
be governed by piston ring design, ring and liner materials, and lubrication. During en-
gine development, or during the running-in process, occasionally engines experience a
catastrophic form of wear known as scuffing. Generally, this is initiated by inadequate
lubrication of the ring and cylinder wall interface, particularly during the critical period
immediately after TDC of the power stroke, where the gas pressures and tempera-

tures are high and the ability to develop protective hydrodynamic films is severely

2
handicapped by low piston speed. This condition causes asperities contact between

the rings and the cylinder wall.

Although there are various methods available for ring wear reduction, the fun-
damental mechanism causing the ring wear is still not completely known because of
the large number of interacting factors. Ring wear is one of the most complicated wear
phenomena since it often involves processes associated with other modes of wear
such as fatigue, abrasion, corrosion as well as adhesion. Generally, the formation of a
ring wear pattern can be accounted for qualitatively by examining the variations of
speed, viscosity, temperature, gas pressure behind the ring, and contact geometry be-
tween the ring and cylinder wall. However, so far there has been no reliable quantita-
tive analysis of a ring wear pattern for an engine under various operating conditions.

Therefore, the first purpose of this research is to understand how friction is de-
veloped for a piston ring, and attempt to evaluate the wear rate that occurs between
the ring and cylinder wall through the use of this ring friction analysis. The second is
to study the wear mechanisms under such low wear rates and investigate the effect of
the power output and engine speed on ring wear to provide a reasonable estimate for
the life of the piston ring. The third is to provide a useful research tool for developing

new materials and lubrication strategies for the piston ring.

1. 2 Dissertation Arrangement

This dissertation is arranged as follows:

Chapter 2 reviews and discusses the piston ring tribology. Experimental stud-
ies on ring wear measurement follow to provide background on the theoretical analy-
sis of ring friction and wear.

The inter-ring gas pressures determined from the mass flow rate passing

through the ring gaps are discussed in Chapter 3. A complete flow chart is presented

3
to provide a detailed description of the gas flow analysis. The ring position is also

studied to find the pressure behind the ring. Simulation results under different engine
operating conditions are presented.

In Chapter 4, assuming axial symmetry throughout the circumference of the
ring, ring friction analysis is performed using the Reynolds equation, the load equa-
tion, and the Stribeck diagram. Depending on system conditions, the ring may demon-
strate different lubricated conditions. Ring friction has been calculated with the lubri-
cated regimes determined from the nominal minimum film thickness, which is defined
as the ratio of the minimum oil film thickness over the surface roughness. The mini-
mum oil film thickness and ring friction force are illustrated at each engine speed.

In Chapter 5, the mathematical model of fire ring wear is presented with the
aid of ring dynamics analysis. Simulation results are presented to support the analyti-
cal findings.

In Chapter 6, some concluding remarks drawn from this research are presented.

In Chapter 7, some limitations of this model, and recommendations for future
research work are discussed.

In the Appendices, the TRIM Simulations [1] which have been conducted for
the Radioactive Ion Implantation (RII) studies are presented along with some prelimi-
nary results. Analysis of measurement sensitivity is discussed for the purpose of pro-
viding estimates of the quantity of the experimental data required to perform in-situ

piston ring wear measurement using the R1] or SLA technique.

CHAPTER 2

BACKGROUND

Reduction of friction between the piston ring assembly and cylinder bore, ei-
ther through improved design, or through the use of more suitable contacting materi-
als, or again through the application of better lubricating substances, is an extremely
important problem for improving fuel economy and energy efficiency in engines. Since
piston ring tribology has been a subject of interest for many automotive researchers
and designers for a long time, numerous papers containing significant and useful con-
cepts, and reviews for research and technologies are available. These topics with pre-
viously published information are reviewed and discussed, and experimental studies
on ring wear measurement are described to provide background on the theoretical

analysis.

2. l Piston Ring Design
The construction of a typical piston and ring assembly is shown in Figure 2.1
[2]. There are two types of rings: compression and oil rings. Compression rings are
designed to seal the clearance between the piston and cylinder to retain gas pressure
and minimize blowby, and oil rings are designed to limit the passage of oil from the
crankcase to the combustion space. Automobile engines normally use two compres-
sion rings and one oil ring, though two-ring designs exist. Most piston rings are made

of gray cast iron because of its excellent wearing properties in all kinds of cylinder
4

5
bores. The axial profiles are chosen to facilitate hydrodynamic lubrication [2]. In to—

day’s gasoline and diesel engines, upper compression rings are normally the barrel
face design, having a smooth face surface for lower initial engine friction, and lower
compression rings are the taper face design. Wear resistant coating is usually applied
on the outer ring surface. In compression rings, the force between the piston ring and
the cylinder wall is due partly to the ring tension and partly to the gas pressure that
leaks into the groove between the ring and the piston [3]. Because of their spring ac-
tion, piston rings press against the cylinder wall at all times, and the gas pressures
behind the rings are given as a function of piston speed and load. To maintain satisfac-
tory levels of oil control and blowby, minimum surface pressure must be maintained

on both compression and oil rings.

Piston crown

Molybdenum-filled
Compression rings

/ Upper compression ring

I .
fi 1 1 Ring belt I

Oi ' . .
1 ring Lower compressron ring

Segment
PShn
Q Chrome-plated
:
E)

\I Expander

Figure 2.1 Schematic of piston and ring assembly in a piston engine [2]

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Oil ring assembly

6
Because of the importance of rings in normal engine operation and the poten-

tial energy savings through improved design, several studies have been conducted
over the years. McCormick et al. [4, 5] presented new developments in piston ring
designs and their relationship to blowby and oil control. Sunden and Schaub [6] dis-
cussed practically orientated principles for the successful manufacture of gray cast
iron piston rings and suggested that, when considered with the sciences of strength of
materials and diesel engineering, the subject of piston rings is almost an embodiment
of the wider subject of tribology. Taylor and Eyre [7] have presented a more detailed
review of piston ring design and piston ring materials as well as cylinder liner materi-
als. Hill and Newman [8] have discussed theoretical considerations affecting piston
ring friction and their implication in ring design. Holt and Murray [9] have observed
that the compression ring of the two ring piston system can be used with advantage
as the top compression ring of an orthodox three ring piston. Young et a1. [10] have
attempted to develop a new ring face coating to extend the life of the internal combus-
tion engine. In recent studies on friction and scuff resistance of ceramic coatings for
rings, Yoshida et a1. [11] revealed that ceramic coatings show lower friction than con-
ventional chromium plate and molybdenum spray coating. Heywood [2] suggested
that the major design parameters which influence ring assembly friction are ring
width, ring face profile, ring tension, ring gap which governs inter-ring gas pressure,
and ring land width. Jeng [12] has also attempted to investigate the effects of engine
design parameters on uibological performance, and suggested that the piston speed,

surface roughness, and ring tension are the most important design variables.

7
2. 2 Piston and Ring Assembly Friction

Energy losses due to friction between the ring assembly and cylinder wall are
purely mechanical energy losses. Most of the heat generated by this friction is dissi-
pated and wasted. In addition to this direct energy loss, excessive wear or scuffing
can result from a high friction force due to poor lubrication. It is well known that even
a small reduction of piston friction results in both better mechanical efficiency and im-
proved fuel economy. Piston friction refers to the friction forces on cylinder surfaces
caused by the reciprocating piston assembly. Measurement of piston frictional forces
during engine operation is valuable for improved fuel economy and good engine de-
sign.

Friction measurements have been conducted by many automotive researchers.
Furuhama et al. [13] canied out piston friction force measurement to obtain the ef-
fects of piston clearance, surface roughness, lubricant, ring size and contact pressure
on the piston friction forces; they suggested that the piston friction is nearly propor-
tional to the square root of the piston speed and the oil viscosity. Furuhama and
Sasaki [14] introduced the movable bore with a pressure balancing device to measure
the friction forces. With the aid of this movable bore method, Ku and Patterson [15]
developed the fixed sleeve method to measure the instantaneous friction of the piston
and ring assembly. The results show that the surface finish of the ring and cylinder
wall affected friction significantly at TDC of the compression stroke. Uras and Patter-
son [16, 17] investigated the effect of the piston variables on piston ring assembly
friction using an instantaneous IMEP method, and observed that during break-in, pis-
ton ring friction gradually reduces throughout the stroke. This suggests that some
mixed or boundary friction exists throughout the stroke. Slone et al. [18] have devel-
oped a new bench tester for the laboratory simulation of the piston ring and cylinder
wear; they observed that a lubricated surface was smoothed considerably when worn,

while a unlubricated surface was further roughened. Simmons er al. [19] have at-

8
tempted to evaluate the fire ring wear as a function of fuel ignition quality. Fire ring

wear was observed to increase rapidly as the ignition quality of the fuel dropped. The
wear rate also increased for a given ignition quality fuel as the injection was advanced
or retarded about the stock timing setting.

In addition to experimental research, numerous theoretical analyses on piston
and ring assembly friction have been undertaken. From experimental data, Bishop
[20] derived empirical equations which describe the magnitude of the most important
factors determining the cycle efficiency and motoring friction of an engine. Ting and
Mayor [21, 22] developed an analytical method for determining the inter-ring gas
pressure, oil film pressure, and piston friction force for a reciprocating piston engine.
This method has been used for predicting the bore wear pattern of actual engines.
McGeehan [23] has presented an excellent review of piston and piston ring friction. It
has been revealed that the piston ring friction is predominantly hydrodynamic with lo-
calized contact between the ring and cylinder wall at TDC firing during transient con-
ditions, and under hydrodynamic conditions ring friction is proportional to the square
root of the viscosity. Rohde [24] developed a ring model which allows the study of
ring friction performance in the mixed lubrication regime. The dependences of the ring
friction and power losses on surface topography, lubrication properties, and engine op-
erating conditions have been investigated. Ting [25] reviewed some recent theoreti-
cal and experimental investigations related to piston ring tribology in the areas of fric-
tion determination, friction reduction, lubrication, and oil consumption. Patton et a].
[26] have developed a friction model which predicts friction mean effective pressure
for spark-ignition engines. This model provides reliable estimates of spark-ignition
engine fmep, and offers a useful tool for understanding how the major engine design
and operating variables affect individual component friction. Yagi et al. [27] suggest-
ed that frictional mean effective pressure is nearly proportional to a non-dimensional

number given by piston stroke, mean equivalent crank diameter, and cylinder bore

9
through the experimental analysis of total engine friction.

The lubrication of the piston skirt is mainly hydrodynamic, while that of the
ring assembly is partly in the mixed lubrication regime [17, 20, 28, 29], which is a
transition state between hydrodynamic and boundary lubrication. In the hydrodynamic
lubrication regime, where the fluid film is thick enough to prevent metallic contact be-
tween ring assembly and cylinder wall, friction is caused only by the viscous shear of
the fluid film. However, as the thickness of the fluid film is reduced to the order of the
surface roughness, metallic contact increases and boundary lubrication becomes domi-

nant [28]. In contrast with hydrodynamic lubrication, long-term boundary lubrication

results in wear.

2. 3 Lubrication and Oil Consumption

The junction between the piston ring and cylinder liner is subjected to excep-
tionally severe dynamic conditions, since the normal load, sliding velocity, and tem-
perature, hence lubricant viscosity, all vary throughout the cycle. In addition, the
shape of the junction might also change throughout the cycle due to elastic and ther-
mal distortion and wear of the piston ring, the piston, and the cylinder liner [30].
Thus, effecrive lubrication has to be achieved with minimum oil consumption.

A number of valuable experimental and theoretical determinations of the oil
film thickness between the piston rings and cylinder liner have been reported. Stewart
and Selby [31] presented a comprehensive review of the published information on the
substantial influence on engine performance of engine oil viscosity. Brown and Hamil-
ton [32] have observed the large negative pressures in oil film lubricating a piston
ring immediately after dead center position, and suggested that these pressures have

the important effect of reducing the minimum film thickness. Dowson et a1. [29] at-

10
tempted to predict the cyclic variation of film thicknesses generated between the ring

and cylinder wall. This report confirmed the view that hydrodynamic lubrication is
prevalent in most of the operating cycle. However, the predicted minimum film thick-
ness, which generally occurs after TDC, indicated that these must be transition to
boundary or mixed lubrication in this region. It has been known that when the engines
are operated on hydrogen as the fuel, all carbon compounds in the exhaust gas
(except for those already in the intake air) can be attributed to the burning of the lubri-
cating oil. By using this principle, Furuhama and Himma [33] could measure the car-
bon dioxide (C02) in exhaust gas, and accurately measure the oil burned. Reipert and
Buchta [34] presented new methods of piston design in relationship to oil consump-
tion and blowby. Boisclair et al. [35] attempted to illustrate the time-dependent ther-
mal environment around the top piston ring and lubricant; they observed that because
of major transient effects, high lubricant temperature is experienced not only at top
ring reversal but also down the liner to bottom ring reversal. By synthesizing recent
technical papers, McGeehan [36] presented an outstanding review on the mechanical
factors affecting oil consumption in four cylinder diesel and gasoline engines and the
practical solutions achieved by design modifications. The progress in understanding
hydrodynamic lubrication of piston rings has not yet led to a satisfactory procedure for
a prediction of oil consumption in an engine, since the net transport of oil past a piston
seal resulting from the presence of the lubricating film is only one component in the oil

consumption mechanism [37]. Further realistic ring lubrication theory is expected to

be deve10ped.

1 1
2. 4 Types of Wear

Adhesive wear affects certain parts of the engine. In the upper cylinder, metal-
to-metal contact between piston, rings, and cylinder walls takes place each time the
engine is started because there is insufficient oil in the top portion of the engine. Oil
with antiwear additives and low viscosity at low temperatures can provide some rem-
edy [2].

Abrasive wear arises from the presence of atmospheric dust and metallic de-
bris from corrosive and adhesive wear in the lubricating oil. Efficient air filtration is es-
sential to eliminate the abrasive particle impurities [2].

Corrosive wear occurs when sliding takes place in a corrosive environment. In
the absence of sliding, the products of corrosion form a film on the surfaces, which
tends to slow down or even arrest the corrosion, but the sliding action wears the film
away, so that the corrosive attack can continue [38]. Corrosive attack by acidic prod-
ucts of combustion is one of the chief causes of cylinder and ring wear. This effect is
worse at low cylinder wall temperature [2]. It has been demonstrated that corrosion
prevention is best accomplished through additive formulation designed to prevent the
accumulation on the metal surface of the precursors to perforrnic acid formation and to
provide excellent dynamic anti-wear characteristics [39]. Nickel coating also has
been proven as an effective and reliable technique to protect pistons from combustion

knock erosion [40].

12
2. 5 Experimental Studies on Ring Wear Measurement

2. 5, 1 Ring Wear Mgasgrement Using SLA

SLA (Surface Layer Activation), also called TLA (Thin Layer Activation) us-
ing radioactive ions has been used for extremely sensitive wear measurements of cer-
tain metallic surfaces. This wear study method, developed over a decade ago [41, 42],
has proven to be a valuable diagnostic tool in evaluating piston ring wear without dis-
assembling the engine [43 - 48]. Schneider and Blossfeld [49], and Konstantinov er
al. [50] studied the rotational movement as well as the wear of piston rings with the
aid of SLA. Particularly, Schneider et al. [51] investigated the effect of speed and
power output on piston ring wear in a diesel engine using SLA. It has been shown
that piston ring wear rates increase significantly with decreasing speed at constant

power output, and it increases with increased power output at all speeds.

2. 5, 2 The Effect of Ion Implantation 9n Friggion and ngr
Shepard and Sub [52] investigated the effects of ion implantation on the fric-
tion and wear behavior of metals and revealed that a significant reduction in friction
and wear of the iron and titanium systems can be attributed to a hard layer formed
during the ion implantation processes. This hard layer minimizes plowing and subsur-

face deformation and hence reduces the delamination wear process. Ion implantation

research suggests that for high doses (>1016 ions/cmz) of heavy ions with low bom-
barding energies (typically several MeV or lower), a variety of material properties
can be changed. These changes include arnorphization, production of defect clusters,
destruction of crystallinity or the production of a new phase, and surface hardness al-
teration. Such ion doses are comparable to the primary proton beam used in SLA, and

are at least several orders of magnitude larger than that needed to produce a useful

level of activity (~lttCi) by direct implantation. The lower mass of the proton beam

13
may not result in the damage absorbed by the MSU research group for the heavy ions

[53].

2 . R i iv Ionlmln inTchni

Mallory et al. [54] have presented the feasibility of implanting a radioactive

ion (e.g., 7Be or 22Na) in ceramic or metallic samples with a dose intensity suitable

for viable wear studies. This Radioactive Ion Implantation (RII) technique could be
applied to study piston ring wear. The activated piston ring emits a low-intensity y-

ray, which penetrates the cylinder wall. This y-ray signature associated with wear

particles is monitored with a NaI or Ge detector. The amount of wear for a given peri-

od of time can be evaluated with the y-ray spectra since the wear is inversely propor-
tional to the measured radiation. This technique can also be used to assess the per-
formance of filtration systems [55].

The analysis of the measurement sensitivity is discussed in the Appendices
for the purpose of providing reasonable estimates of the implantations needed to ob-
tain in situ piston ring wear measurement using the R11 technique. The polyenergetic
TRIM simulation has been used to find an optimized energy of the implanting beam
and a suitable set of absorbers necessary to produce a desired uniform dose-depth
profile in the surface of a sample for RII studies.

Further detailed descriptions on the TRIM simulation and R11 including mea-

surement procedure can be found in Appendices

CHAPTER 3

GAS FLOW ANALYSIS

3. 1 Introduction

Small volumes formed at interfaces to connecting parts in an engine’s combus-
tion chamber are called the crevices. Gas flows into and out of these volumes during
the engine operating cycle as the cylinder pressure changes. Total crevice volume is a
few percent of the clearance volume, and the piston and ring crevices are the dominant
contributors. When the engine is warmed up, dimensions including crevice volumes
are changed.

The crevice processes occurring during the engine cycle are described as fol-
lows [2]: As the cylinder pressure rises during compression, the unburned mixture or
air is forced into each crevice region. Since these crevices are thin, they have a large
surface to volume ratio; the gas flowing into the crevice cools by heat transfer to close
to the wall temperature. During combustion while the pressure continues to rise, the
unburned mixture or air, depending on engine type, continues to flow into these crev-
ice volumes. After the flame arrives at the crevice entrance, burned gases will flow in-
to each crevice until the cylinder pressure starts to decrease. Once the crevice gas
pressure is higher than the cylinder pressure, gas flows back from each crevice into
the cylinder. The back-flow of gases into the combustion chamber can lead to exces-
sive exhaust hydrocarbon and poor economy. For these reasons and others, under-
standing the processes involved in piston ring sealing are critical to a good engine de-

sign.

l4

l 5
Head

Combustion Chamber

 

 

 

 

 

 

 

 

 

 

Top Land Crevice
Ring Groove Clearance 1
_..l ‘—
'. ................. ’ ..... Ring Side Clearance (h)
1-2 I
Region behind Rings 2
2-3 N
3
3-4 J
Oil Ring Assembly l 4
Crankcase

    
  

.3»..-

TOP Compression Ring Gap ...»~ ..
. mw-\

 

Figure 3.1 Detailed schematic of clearance volume in a piston engine

16
The volumes between the piston, piston rings, and cylinder walls are shown

schematically in Figure 3.1. These crevices consist of a series of volumes (numbered
1, 1-2 etc.) connected by flow restrictions such as ring side clearance and ring gap.
The geometry changes as each ring moves up and down in its ring groove, sealing ei-
ther at the top or bottom ring surface. The gas flow, pressure distribution, and ring
motion are therefore coupled [2]. The pressure distributions and ring motion can be
determined by analyzing these crevices as volumes connected by passage ways, with
a prescribed cylinder pressure versus crank angle profile coupled with a dynamic mod-
el for ring motion, and assuming that the gas temperature equals the wall temperature
[56, 57].

The gas that flows from the combustion chamber past the piston rings into the
crankcase is called the blowby. If there is good contact between the compression
rings and the bore, and between the rings and the bottom of the grooves, then the
only leakage path of consequence is the ring gap. Blowby of gases from the cylinder
into the crankcase removes gas from these crevice regions and thereby prevents
some of crevice gases from returning to the cylinder [2]. However, if the blowby is ex-
cessive, compression is reduced, emissions increase, and the lubricating film may de-
teriorate, affecting engine reliability and life.

Gas flow analysis has to be performed prior to the ring friction analysis since
the gas pressures above and below a ring are used as boundary conditions in the cal-
culation of the oil film pressures, and gas pressures behind the ring directly affect the
ring friction force deve10ped between the ring and cylinder bore. Ting and Mayor [21,
22] developed a method for computing the inter-ring gas pressure variations through-
out the engine cycle. This analysis is reviewed and discussed with the assumptions

and limitations of the theory.

l7

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Combustion Chamber
P1 921
T1 dt

4 V2 1 A1 [ Upper Compression Ring
P2 dmz m1
T2 '21? p,
V1

J v3 A2 [ Lower Compression Ring
P3 r112
r3 2‘: 92
V2 dt

] V4 1 A3 F Oil Ring Assembly
P4 = atm.
T4
Crank Case

Figure 3.2 Orifice Volume Model of Ring Pack

18
3. 2 The Mass Flow Rate

The ring gap areas have been considered as the most important factors in de-
termining the mass flow rate in an engine’s crevices [21, 22]. The size of the ring
gaps is increased due to the combined effects of ring and liner wear. Hence, the effect
of ring gap area on the mass flow rate is critical in the ring design. The mass flow rate
through each ring gap needs to be determined to calculate the inter-ring gas pressure.
An orifice volume model of ring pack is shown in Figure 3.2. The volumes represent
the inter-ring spaces that are formed by two adjacent rings, the intervening piston

land, and the cylinder bore.

q

 

1w
\ [I

unit mass

unit mass
\ /

 

 

 

 

 

\ I
—’ V1 V2
P}: p], “1 P2’ p2’ “2

 

 

 

 

 

Figure 3.3 Schematic illustration of energy balance [28]

19
The following assumptions were made to model mass flow rate [21, 22]:

1) The ring gaps are the only gas leakage paths.

2) The rate of heat transfer is small, such that the gas flow through the ring pack is an
unsteady adiabatic flow satisfying the perfect gas law.

3) The orifices are assumed to be equal in area to the effective leakage path formed
by the ring gaps and the radial clearance between the piston and cylinder bore.

4) The gas flow is considered as a one-dimensional flow through the orifice with a
constant discharge coefficient.

5) The friction effects are small such that the flow is isentropic.

6) The combustion chamber pressure remains unaffected despite the gas leakage.

7) The crankcase is considered to be at atmospheric pressure.

Consider a gas passing through a control volume as shown in Figure 3.3.
where W = the work done on or by the gas

q = the heat added per unit mass

v = velocity

u = internal energy per unit mass

P = pressure

p = gas density

and subscripts 1 and 2 denote each condition at the inlet and the exit, respectively.

For a unit mass, neglecting the potential energy changes, the Energy Equation is giv-

en as

q+P1/p1—P2/p2+W=u2-u1+(v22—v12)/2 (3.1)

Because adiabatic flow has been assumed, q = 0. If, in addition, W = 0, then Equation
(3.1) reduces to

20
P1 /p1 - 1>2 /p2 = u2 — u1+(v22 — v12)/ 2 (3.2)
For an ideal gas, the following relation is valid:

P = pRT (3.3)

where T is the absolute temperature and R is the ideal gas constant.
The specific heat of a gas is the amount of heat required to change the temper-
ature of a unit quantity of the gas one degree. For an ideal gas, the specific heat ratio

is given by

y = Cp / Cv (3.4)

where Cp denotes the specific heat at constant pressure and Cv is the specific heat at

the constant volume. Since the change in internal energy for a non flow constant vol-

ume process depends only on the temperature difference
u2 — u1 = Cv('l'2 — T1) (3.5)
For a constant pressure process
Cp = C" + R (3.6)

By combining Equations (3.4) and (3.6)

Cp = R /(r—1) (3.7)
Cv = 7R /(Y-1) (3.8)

21

Therefore, for an isentropic, adiabatic process, the following relationship is valid:

P, /P2=(p, /p2)7 = (fl/rpm“) (3.9)

For a steady one dimensional gas flow, mass flow rate is given as

dm /dt = pAv (3.10)

where m = the mass of gas
A = a cross-sectional area of flow

v = the velocity of gas

Then the mass flow rate passing through the upper compression ring gap as a function

of crank angle can be expressed as

am1 /d9 = (dmlldt) /(d9/dt)

= 92 A1V2 “0 (3.11)

Since v1=0, the velocity of the gas (v2) through the ring gap can be found by substi-

tuting Equations (3.5) and (3.3) into Equation (3.2).

v2 = {2 [P1 /pl - P2 /p2 + C,(T,— r2)]}1/2
= {2 [RCFI- T2) + RCTI- T2) /(7 - 1)]1"2

= [2 vRTla— Tzrrl) /(7- mm (3.12)

22
Again after substituting Equation (3.9) into Equation (3.12), the gas velocity becomes

v2 = {2 YRTlll - (P2/P,)‘H>”1/(v— 1))"2 (3.13)

Accordingly, the mass flow rate passing through the top ring gap is given by

am1 /d0 = A1 /co{2 yRTlpzzll — (P2 /P,)(l'1W]let-1))"2

= A, /co{2 Y/(Y-1)'(Pzlpl)2'(PIP1)'[1 -—(1>2 Harm/'11 1'2 (3.14)

Substituting Equation (3.9) into Equation (3.14), and using Equation (3.3) the mass

flow rate for P1 > P2 is determined as

am, /d6 = A, m2 y/(y— mp, /P1)m-P12/(RT1)'[1 -(1>2/1>,)‘*‘1>"'11”2

= A,K,-(P,/ rye-(P, /P,)‘”-[l - (1>,/1>,)(*‘1>"1“2 (3.15)

where K, a (2 y/[R(y- 1)]]1’2/(0
If Pl < P2

am, me = -- z>t,I<,-(P2/T,m)-(P1 mal/{[1 - (Pl IP2)(I"1)”11’2 (3.16)

where the negative sign signifies the reverse gas flow.

Similarly, for P2 > P3

drr12/d6 = A2 Kl-(P2 / T21”)-(P3 IP2)1/7-[1 - (P3 1P2)(Y'1)/Y]1/2 (3.17)

23

If P2 < P3

dm2 ld9 = - A2 Kl-(P3 / T31”)-(P2 /P3)1”-[1 -— (P2 IP3)(H)”]"2 (3.18)
Also, for P3 > P4

am3 /d9 = A3 Kl-(P3 /T31’2)-(P4 IP3)1”-[l — (P4 IP3)(H)”]1’2 (3.19)
If P3 < P4

am, me = - A, K,-(P, / T41/2).(p3 mam-[1 — (P3 /P.)‘H>”1 1'2 (3.20)

However, if the pressure downstream of the ring gap is less than the critical
pressure, the pressure in the ring gap is always the critical pressure and the mass
flow always equals the maximum value; whereas if the pressure downstream of the
ring gap is greater than the critical pressure, the mass flow rate is determined by one
of equations above according to the conditions [21, 22]. The critical pressure, where
the mass flow rate reaches a maximum value, can be determined by differentiating

this given equation with respect to Pi (where i = 2, 3, 4) and setting this result equal
to zero. Consider Equation (3.15) as an example.

Defining P2 /P1 5 x,

dm1/d0 A1K1'(P1 /T11/2).x1/Y.[1 _ x(‘t'-1)/‘t']1/2

AlKl-(Pl /r,1fl).[x?fl— FWD/'11” (3.21)

Differentiating Equation (3.21) with respect to x and setting the result equal to zero

24
gives

1 f2.[x2/Y_ x(t‘+l)/Y]-l/2.[2 H.x2/Y-l_(y+1)/Yx1/Y] = 0

Since [xm - XWHVH’I’Z at 0 for Cp at Cv

ZH-x(“”"-(Y+1)/Y=0

Thus

x =1<v+1)/217/(H>

That is, the critical pressure is found to be

Pc = P,[( 7+1) / 217“”) (3.22)

For combustion gases passing through the ring pack, y is assumed to be 1.3 and the

critical pressure is 0.546 Pi (i = 1, 2, 3, 4) [21, 22].

Therefore, if P2 S 0.546 P1, Equation (3.15) becomes
dml /d6 = AIKIK2 -(P1 ITIVZ) (3.23)

where K15 {2 Y/[R(y— 1)]}“2/0)

K2 5 (0.546)1”-[1 — (0.546)(Y‘1)/Y]1/2

For Pl <P2 , if P1 5 0.546 P2

25

am, /d9 = -A1 1(le -(P2 / T21”) (3.24)

Similarly for P2 > P3 , if P3 .<. 0.546 P2

dm2 /d9 = A2 K1K2-(P2 / r21”) (3.25)

For P2 < P3 , if P2 5 0.546 P3

(rm2 /d9 = —A2 Kle-(P3 / r31”) (3.26)

For P3 > P4 , if P4 S 0.546 P3

dm3 /d9 = A3K1K2 .(P3 n31”) (3.27)

For P3 < P4 , if P3 S 0.546 P4

dm3 /d9 = —A3K1K2 -(P4 / r41”) (3.28)

However, the actual mass flow rate through an orifice is less than the theoreti-
cal mass flow rates given by Equations (3.15) to (3.20) and Equations (3.23) to
(3.28) because of the friction losses of the gas passing through the ring gap and the
convergence of the gas streamlines as they pass through the ring gap [21, 22]. The

actual mass flow rate through a ring gap is given by

26

(drn mom = Kc(dm momma“, (3.29)

where Kc = orifice discharge coefficient

An orifice discharge coefficient of 0.65 has been taken for the calculation of inter-ring
gas pressure since the ring gap was considered as a square-edge orifice [21, 22].

3. 3 The Determination of Inter-ring Gas Pressure
The mass of the gas bounded in the volume between the top and second ring

shown in Figure 3.2 is given by

m1 = plv1 (3.30)

Substituting Equation (3.3) into Equation (3.30)

P2(9) = m1(6)RT2 WI (3.31)

At the crank angle of 0+A0, the pressure in the volume is given by

P2+AP2 = er/v1 {1111(9) + ram, /d9—dm2 meme) (3.32)

Hence, the rate of pressure change is written as

dP2 /de z RT2/V1(dm1 /d9 - am, MB) (3.33)

27
Similarly

dP3 me ~ RT3/V2(dm2 /de - cim3 /d9) (3.34)

Therefore, the inter-ring gas pressure at each crank angle can be determined by

marching through the cycle.
P2(6+A6) a P2(O) + (d9 l2)[(dP2/d9)e—(dP2/dO)O+A9] (3.35)
P3(6+A9) = P3(9) + (d9 l2)[(dP3/d9)9—(dP3/d9)3+Ae] (3.36)

The convergence criteria for the calculation of the inter-ring gas pressure is
critical. The iterations are continued until the convergent gas pressures are obtained.

Since the gas flow, pressure distribution, and ring motion are coupled as stat-
ed earlier, the axial movement of a piston ring in the ring groove needs to be consid-

ered to determine the pressure behind the top compression ring (P12)-

3. 4 The Axial Motion of the Fire Ring in the Groove
The axial motion of the ring can be predicted by using a simple force balance
model [56, 57] with the same assumptions used in the evaluation of inter-ring gas

pressures as follows:

The force due to the gas pressure acting on the ring is expressed as

W:
Fp = no, I0 P(x)dx (3.37)

28
where Dr = diameter of the ring

Wr = width of the ring

x = coordinate in radial direction

P(x)=pressure distribution on the ring

Again assuming axial symmetry throughout the ring circumference, i.e., neglecting the
twist, the rotation of the ring, and cylinder bore distortion; the pressure distribution on
the ring can be found by considering the pressure below and above the ring with the

substantial pressure drop across the ring.

1’00 = [P1(9)+P1-2(9)l /2 - [P1_2(9)+P2(9)] /2

= [P1(9)-P2(9)] /2 (3.38)

Equation (3.37) becomes

Fp z nDrWr[P1(9)-P2(9)] /2 (3.39)

While the inertia force is given by

Fi = mrap (3.40)

where ap denotes the acceleration of the piston. The direction of this force is depen-

dent on the direction of piston motion. The friction force on the cylinder wall is usually
neglected in the determination of the ring position in the groove since it is relatively
small compared to the inertia and pressure force [28]. Also unlike Kuo et al. [57], an

oil resistance between the ring and the groove is not included since there is too much

29
uncertainty concerning the presence and coverage of the oil film [5 8]. Therefore, the

equation of the motion for the ring is found to be

nDrW,[P1(9)—P2(9)] /2 + mrap = m,(d2h/dt2) (3.41)

where h represents the ring side clearance. As long as the sum of the forces is posi-
tive, the ring seats in the bottom of the groove. But as the sum becomes negative, the
ring moves upwards and finally settles on the top side of the groove. Equation (3.41)
can be solved by applying the fourth order Runge-Kutta integration method. Thus, the
pressure behind the fire ring can be determined considering the ring position in the

groove.

3. 5 Results and Discussion

The specifications of the diesel engine used in this study are shown in Table 1.
As a completed data set was not available, it has been pieced together based on
available data for similar engine. The combustion gas pressures have been generated
to give the same power output at each engine speed. Further detailed descriptions on
IMEP and the power output follow in Section 4.2.

The procedure for the determination of inter-ring gas pressures at each engine
speed is shown in Figure 3.4. The computer simulation has been performed using a
ring dynamics analysis program, which has been under development for the past five
years [28]. Figures 3.5 through 3.8 depict the gas pressures in each region deter-
mined from this analysis. The fire ring motion in the groove at each engine speed is
shown in Figures 3.9 through 3.12. Comparing these Figures - Figures 3.5 and 3.9,
Figures 3.6 and 3.10, Figures 3.7 and 3.11, and Figures 3.8 and 3.12 - it is found that

30
the gas pressure behind the ring depicted as the short-dotted line in each figure is

definitely influenced by the ring motion in the groove. In other words, if the ring set-
tles on the top side of the groove, the gas pressure behind the fire ring is the same as
the gas pressure between the fire ring and lower compression ring, whereas if the ring

seats at the bottom of the groove, it becomes the same as the combustion pressure.

31

 

 

Table 1 Specifications of the Engine
Engincflata
Engine Type : Diesel Engine
Engine Displacement: 12.7 L
Tested Engine Speed : 1200 rpm, 1500 rpm, 1800 rpm, 2100 rpm
Bore Diameter : 130 mm

Bore Surface Roughness : 0.15 pm
Stroke : 160 mm

Connecting Rod Length : 264.8 mm
Connecting Rod Weight : 4.704 kg

Engine Egg Temperatum

9 mm from the Deck Top: 159 °c
22 mm from the Deck Top : 124 0C

165 mm from the Deck Top: 113 °c

Temperature Data
Mean Combustion Temperature : 326 0C
Upper Compression Ring Groove Temperature : 210 0C
Lower Compression Ring Groove Temperature : 177 0C
Oil Ring Groove Temperature : 135 0C
Second Land Temperature : 191 0C
Third Land Temperature : 149 °C

Sump Temperature : 121 °C

 

 

32

 

 

Table l (Cont’d)

21mm

Piston Diameter : 129.9 mm

Piston Height: 157 mm

Distance from the Deck to TDC: 1.5 mm
Piston Pin Weight : 1.828 kg

W

SAE30

W

Rin

Friction Coefficient : 0.1

The Nominal Minimum Oil Film Thickness for Mixed Lubrication (A) : 1 s A S 5
Da

Upper Compression Ring Gap Area : 0.277 mm2
Lower Compression Ring Gap Area : 0.245 mm2
Oil Ring Gap Area : 0.865 mm2

Upper Compression Ring Groove Area : 96 mm2
Lower Compression Ring Groove Area : 105 mm2

Oil Ring Groove Area : 147 mm2

Land gs] ngve Volume Data

Upper Compression Ring Groove Volume : 991 mm3
Lower Compression Ring Groove Volume : 952 mm3
Oil Ring Groove Volume : 721 mm3

Second Land Volume : 4180 mm3

Third Land Volume : 689 mm3

 

 

33

 

 

Table l (Cont’d)
II C . B. D
Ring Thickness: 3.91 mm
Ring Width : 5.20 mm
Ring Diarnetral Tension : 51.59 N
Ring Weight : 0.045 kg
Ring Location : 29.8 m
Surface Roughness : 0.127 um
Groove Gap Width : 0.244 mm
Ring Face Type : Parabolic
Crown Height : 0.014 mm
Crown Offset : 0
Lowg Compression Ring
Ring Thickness : 2.84 mm
Ring Width : 5.18 mm
Ring Diameu'al Tension : 44.48 N
Ring Weight : 0.035 kg
Ring Location : 47.4 m
Surface Roughness : 0.127 um
Groove Gap Width : 0.264 mm
Ring Face Type : Parabolic
Crown Height : 0.097 mm
Crown Offset : 1.143 mm
Oil Ring Data
Ring Thickness : 2.84 mm
Ring Width : 5.18 mm

 

 

34

 

 

Table 1 (Cont’d)
Ring Diametral Tension : 48.93 N
Ring Weight : 0.035 kg
Ring Location : 54.6 m
Surface Roughness : 0.127 um
Groove Gap Width : 0.373 mm
Ring Face Type : Parabolic
Crown Height : 0.082 mm
Crown Offset : 0.914 mm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

no
st '546P1 yes P1> P2 P1554615
. flowrate(l) ————" flow rate (2) .
Ym yes
Ll , A
no 7| flow rate (3) flow rate (4) ~ no
yes I “0
P3554613 P2> P3 P25 .546P3
* flowrate(5) __ .____, flow rate(6) w
Yes Yes
= flow rate (7) . flow rate (8) 2
no no
W
es no
P45 .546P3 y P3> P4 P35 .546P4
flowrate9 . I (flowrate 107
( ) ———1 a ( ) yes
no > flow rate (11) -— _. flowrate(12) :no
v I! II v

 

 

 

 

 

 

[ Calculate Rate of Pressure Rise }:
I
l Calculate New Value of Inter-RingGas Pressure ]

ll

/

Convergence

 

 

no
yes
[ Flow Rate at Next Crank an 1e = Previous Flow Rate ]

 

 

 

no

STOP

 

Figure 3.4 Procedure for the determination of inter-ring gas pressures

Cos Pressure(Kpo)

Cos Pressure(Kpo)

10000

8000

6000

4000

2000

10000

8000

6000

4000

2000

0 200 400 600 800

O 200 400 600 800

 

 

 

 

 

 

Crank ongle(degree)

Figure 3.5 Gas pressure in each region at an engine speed of 1200 rpm

 

 

 

 

 

 

Crank ongle(degree)

Figure 3.6 Gas pressure in each region at an engine speed of 1500 rpm

Cos Pressare(Kpo)

Gas Pressure(Kpa)

37

 

 

 

 

 

 

 

 

 

 

 

 

 

10000 T I T I Y T T I I I' T T
_ — Pl -
" ........ p ‘
8000 — "2 —
: P: :
6000 -— _
)- -I
4000 ~ —
)- -1
2000 — ..
' 'I
0 “b T . 1
0 200 400 600 800
Crank angle(degree)
Figure 3.7 Gas pressure in each region at an engine speed of 1800 rpm
10000 ' ' * l . . . I . l . .
_ Pl ..
_ ........ p ‘
8000 — 1‘2 -
l- - '" - P2 q
6000 — -
4000 -— _.
2000 ~— ~
I— .4
O E:— L . i 1
0 200 400 600 800

Crank angle(degree)

Figure 3.8 Gas pressure in each region at an engine speed of 2100 rpm

Ring Position in the Groove(m)

Ring Position in the Groove(m)

2.5x10-4 T 1* I l
_ I j
r I 7
2.0x10-4 _ J
I- a
I' -4
1.5x10‘4 -— _
P :
I" -
10x10“4 — L
_ ’ I .
5.0x10 5 I I a
e I :
r I -
O l , 1 It 4 1 11 1 I 1 L A
o 200 400 600 800

2.5x10-4 I I ‘ I
\ I .I
_ I 4
2.0x10'4 — 1 —
_ I _
T d
1.5x10‘4 - I —
)- 3 .
r- ; —
h _: ~
I- I -1
_ I
1.0x10 4— I —1
Pl. d
t. I ..
50x10":3 — I I ‘1
L. I I .1
I . ..
O L L I i I l I] l I l l L
O ICC 400 600 800

38

 

 

 

 

 

 

 

 

Crankangle(deg ree)

Figure 3.9 Ring position in the groove at an engine speed of 1200 rpm

 

 

 

 

 

 

 

 

Crankanqle(degree)

Figure 3.10 Ring position in the groove at an engine speed of 1500 rpm

Ring Position in the Groove(m)

Ring Position in the Groove(m)

 

 

T l 1 I r r r
r— 4
4__ -—
I- d
' 1
.- _I
4__ ‘—
5 _ _
O A l L I 1 I - l r I I l L
0 200 400 600 800

39

 

 

 

 

 

 

 

 

Crankangle(degree)

Figure 3.11 Ring position in the groove at an engine speed of 1800 rpm

 

 

 

 

 

 

 

f W I I I r
.1
q
" -I
' 1
)— -1
_ I 4
t
)— I f ‘4
._ [I .4
’- d
4 L i A! L I l
0 200 400 600 800

CrankangIe(degree)

Figure 3.12 Ring position in the groove at an engine speed of 2100 rpm

CHAPTER 4

THE ANALYSIS OF FIRE RING FRICTION

The geometry of a piston in the cylinder bore and the velocity of a piston affect
the lubrication conditions of the ring. Moreover, ring friction might be changed accord-

ing to the engine speed and power output. These topics are reviewed prior to the ring

friction analysis.

4. l Piston Kinematics
The basic geometries of a reciprocating engine are defined by the ratio of cylin-

der bore to piston stroke (Rbk); ratio of connecting rod length to a crank radius (Ru);
and the compression ratio (rc), which represents the ratio of maximum cylinder vol-
ume (Vc+Vd) to minimum cylinder volume (Vc) shown in Figure 4.1. In addition, the

stroke and crank radius are related by

K =2a (4.1)

Typical values of these parameters are given as follows [2]:

1) rc = 8 to 12 for SI (Spark-Ignition) engines, and rc =12 to 24 for CI (Compression-
Ignition) engines.

2) Rbk = 0.8 to 1.2 for small- and medium-size engines, decreasing to about 0.5 for
large slow-speed CI engines.

3) RLa = 3 to 4 for small and medium-size engines, increasing to 5 to 9 for large slow-

speed CI engines.

40

41

 

——-—- ———————— — rc

 

 

 

 

 

 

 

TC

BC

Figure 4.1 The geometry of a reciprocating piston engine, where B = bore, K = stroke,

L = connecting rod length, a = crank radius, 0 = crank angle[2]

42
The location of the piston depicted in Figure 4.1 can be determined as a func-

tion of crank angle as follows :

3(0) = L + a (1 - eose) — (L2 — a2 sin20)m (4.2)

where the crank angle is defined as

6 = wt (4.3)

with the angular velocity, (1), obtained from

a) = 27tN /60 (4.4)

where N is the rotational speed of the crank shaft in rev/min. Therefore, for one cycle

of four-stroke engine operation

0 S t S 41t/0) (4.5)

Thus, the mean piston speed is expressed as

Umean = K / (0 /m)
= KN B0 (4.6)

The instantaneous piston velocity is given by

Up(t) = [dS(0)/d0](d9 /dt)

= a)[a sine + a2sin0 cosO(L2 - azsin20)'m]

43
= a)[a sintot + (a2/2)sin 2mt (L2 - azsinztotym] (4.7)

Differentiating Equation (4.7), the piston acceleration is found to be

‘50) = (02a cos cut + afiazcoszwt -(L2 - azsiantyl/Z
— ofiazsinzmt -(L2 - a2sin2mt)-1/2
+ mazsin (0t cos a)t(—-1/2)-(L2 - a2 sinzmt)'3’2«(—2 azsin (at cos tot-(o)
= (0221 cos (at + (ozazcos 2cm ~(L2 - azsinzmtylfl

+ m2a4sin2mt coszart-(L2 - azsinzart)‘3’2 (4.8)

4. 2 The Indicated Mean Effective Pressure and Brake Power
One of the most significant indicators of the performance of internal combus-
tion engine is the IMEP (Indicated Mean Effective Pressure). IMEP is defined as an
average work delivered at the piston face during a complete engine cycle.
From a thermodynamic analysis, the work performed on or by a chemical sys-

tem during a volumetric change in state is
w =J‘ P dV (4.9)

where W = work
P = cylinder pressure

V = cylinder volume

IMEP is obtained by normalizing this indicated work with the total cylinder displace-

ment volume [59].

44

IMEP: (1ND)! PdV

= (1ND) [04% (dV/d0)d8 (4.10)

where VD denotes the total cylinder displacement volume. Brake mean effective pres-

sure is determined as
BMEP=ne- IMEP (4.11)

where 1],, denotes a mechanical efficiency. A mechanical efficiency of 90% has been as-

sumed in this analysis. Therefore, the brake power is obtained from
BP = BMEP 'VD'Nmu /nll (4.12)

where N"m represents the maxrmum rated engine speed obtained for a maxrmum
mean piston speed, and nR is the number of crank revolutions for each power stroke

per cylinder (two for four-stroke cycles, one for two-stroke cycles). These definitions
are used with the data available in the literature to develop pressure-crank angle in-

formation for the conditions at which the power output is known.

45
4. 3 Reynolds Equation

The oil film thickness developed between the ring assembly and cylinder bore
needs to be determined in order to predict oil consumption and friction loss. However,
the minimum oil film thickness cannot be determined easily since it can be taken as a
function of the surface roughness [28]. It has been observed that the ring can tilt, ro-
tate, or move axially up and down in the groove throughout the engine cycle [49, 56,
57, 60, 61]. However, here the ring motion in the groove is neglected since it is negli-
gibly small compared to piston motion, and thus has little effect on the ring friction
[62].

Therefore, for two dimensional flow of an incompressible lubricant, the Navier-

Stokes equation for the liquid film motion reduces to a Reynolds equation of the form
[2, 29, 30, 63]

8/8x{(h3/ll)-(3P/8x)l + 3/3y[(h3/ll)-(3P/8y)l = -6U,(ah/ax) +12(ah/at) (4.13)

where h = oil film thickness
[.1 = oil viscosity
p = oil film pressure
Up= piston velocity
x = coordinate in the axial direction
y = coordinate in the circumferential direction

t=time

The following assumptions were made to model hydrodynamic lubrication for
the piston ring [21, 22, 28]:
1) The oil film between the ring and cylinder bore is sufficient for hydrodynamic lubri-

cation, thus body forces are ignored.

46
2) The oil viscosity does not change around the ring face but may change with temper-

ature at different positions along the bore.
3) The lubricant is Newtonian and incompressible.
4) The flow is laminar.
5) There is no slip at the boundaries.
The lubrication of piston rings in firing engines is periodic with a period of 41th:)

for a four-stroke engine. Hence

P0) = P(t+41t/€0)

h(t) = h(t+41t/(1)) (4.14)

If axial symmetry is assumed such that there is no flow of lubricant in the circumferen-

tial direction between the ring and cylinder bore, and the lubricant is assumed to be an

incompressible fluid having a mean viscosity ()1) throughout the oil film at any speci-

fied crank angle, then Reynolds equation can be expressed as
a/ax[h3(3p/8x)] = -6uUp(ah/ax) +12u.(8h/3t) (4.15)

A first integration of Equation (4.15) leads to an expression for the axial pressure

gradient in the lubricant film
dp/dx = —6p.UP/h2 + 12tt(x/h3)(ah/at) + C1/h3 (4.16)
where Cl is the integration constant. Therefore, the pressure distribution of the lubri-

cant between the ring and cylinder bore can be found by numerically integrating Equa-

tion (4.16).

47

p = —6uUp11 + numb/8012 + C113 + (‘1 (4,17)

where 11 = Idx/hz
I2 = Ix] h3dx

13=Idxlh3

C2 = an integration constant

The geometry of lubricated junction between a piston ring and cylinder bore is shown

in Figure 4.2. Hence, the boundary conditions are given by

p = P1 at x=0

p = P3 at x=Tr (4.18)

where P1 and P3 represent the pressure above and below the top ring, respectively,
and Tr is the thickness of the ring. These conditions enable the integration constants

Cl and C2 to be determined.

Since dy = me, the normal load on the top ring is expressed as

21! T,
P“ =J' J (P2 + Pm)rdxd0 (4.19)
0 0

where P2 denotes the pressure behind the top ring, Pten represents the pressure due

to the radial tension of the ring, and r is the outside radius of the ring. Since axial sym-

48
merry has been assumed, the load equation is then written as

JTRP + P )rdx = IT'p(x)rdx (4.20)
o 2 “m 0

where p(x) denotes the axial pressure distribution between the ring and the cylinder
bore shown in Equation (4.17). Along with the geometry of the ring face profile, the
cylinder bore temperatures at the ring locations for any crank angle has to be defined
since the oil viscosity is greatly dependent on the temperature. By solving Equation
(4.20) with Equation (4.17), oil film thickness and oil film pressure distribution can be
determined. Further detailed discussion on the determination of the viscosity is fol-

lowed in the next section.

49

Combustion Chamber x

 
  
 
 

Cylinder Liner

Piston Ring

Piston

Lubricant

Region between the Ring

Figure 4.2 Geometry of lubricated junction between a piston ring and cylinder bore

50
4. 4 Lubrication

The lubricant between the ring and the cylinder bore reduces the frictional re-
sistance of the engine to a minimum to ensure maximum mechanical efficiency, pro-
tects the ring and the bore against the wear, and also contributes to cooling the piston
and the cylinder where the friction work is dissipated.

The temperature of the oil and engine parts it contacts, the presence of oxy-
gen, the characteristics of the metal surfaces and debris, and the products of the fuel
combustion influence the oxidation of the hydrocarbon components in lubricating oil

[2]. Of these, high temperature is the primary factor. The top ring groove, where the

temperature easily reaches 250°C is the most critical region. The lubricating oil under
these conditions contributes to deposit formation. These deposits eventually lead to
ring sticking, which results in excessive blowby [2].

The viscosity, which is the most important property of a lubricant is a function
of temperature and pressure [28]. Generally, both friction and film thickness increase
with increasing viscosity under normal engine operating conditions. The density of oil
is little affected unless turbulent motion exists between the ring and the cylinder
bore. The effect of temperature on the oil viscosity is much greater than its effect on
any other physical properties. The viscosity of lubricating oils decreases with increas-
ing temperature. Following is the one of the most frequently used viscosity and tem-

perature relations.
11 = a-exp[b/(T+c)] (4.21)
where a is the viscosity at T = 00. The constants in this equation have been deter-

mined from the known viscosity values at the specific temperature. They are shown in

Table 2. The unit of viscosity is N-sec/mz.

51

Table 2 Constants in the Vogel Viscosity Equation

 

 

 

 

 

 

 

 

 

 

 

 

Vogel Constants
SAE Grade a b c
5 W 0.05567 900.0 110.8
10 W 0.04082 1066.0 116.5
15 W 0.06681 902.0 100.2
20 W 0.02370 1361.0 123.3
20 0.04987 1028.0 108.0
30 0.02370 1361.0 123.3
40 0.07227 1396.0 121.7
50 0.01963 1518.0 122.6

 

 

 

 

4. 5 Fire Ring Friction

The ring may undergo the different lubricated conditions depending on the ef-

fective oil film thickness and joint effective surface roughness between the ring and

the cylinder bore.

The coefficient of friction can be expressed as [2]
f0 = (rt-fb + (l—or)fh (4.22)

where fb denotes the boundary friction coefficient of the metal-to-metal contact, fh is

the hydrodynamic friction coefficient, and (1 represents the metal-to-metal contact

constant varying between 0 and 1. As (rt—)1, the lubrication regime approaches the

52
boundary regime, while as a—)0, it approaches the hydrodynamic regime. Here the lu-

bricant film is sufficiently thick to separate completely the surfaces in relative motion.
The mixed lubrication regime can be found where the transition between these re-
gimes occurs. The lubrication regimes are determined from the Stribeck diagram [6,

23, 24] shown in Figure 4.3 by using the ratio of minimum oil film thickness to joint ef-

fective surface roughness.

10°-

Boundary -._-.,..'Mixed Hydrodynamic

10'1 "

10-2 -

Cocfficient of Friction

 

A, A, t

Figure 4.3 Stribeck diagram for a piston ring

(A = nominal minimum film thickness)

53
4, 5, l Hydepmig Lpprigtipp

If the ratio of effective oil film thickness to the joint effective surface roughness
is large enough, then friction is caused by the viscous shear of the lubricant and the
pressure gradient through the lubricant. Since the lubricant is assumed to behave as a
Newtonian fluid, the shear stress is proportional to the velocity gradient across the

film. Therefore, the hydrodynamic friction force per unit area of the sliding surface in
the axial direction is defined as [28]

fx = h /2(dp/dx) + uUp/h (4.23)

where dp /dx = pressure gradient in the axial coordinate
x = coordinate in the axial direction
h = oil film thickness
[.1 = oil viscosity

Up: piston speed

The evaluation of the oil film thickness and pressure gradient in the axial coor-
dinate have been discussed already in Section 4.3. Here the hydrodynamic friction

force can be determined by integrating Equation (4.23) over the nominal contact area

between the ring and cylinder bore [28].

21! T,
Ff=JI I r, rdxd0 (4.24)
0 0

where Tr denotes the thickness of the fire ring, and r is the radius of the ring. Since

the axial symmeu'y throughout the circumference of the ring has been assumed, Equa-

tion (4.24) can be written as

54

r.
P, = 2n r, rdx (4.25)
0

Again, the geometry of the ring face profile needs to be defined in order to perform this

integration numerically.

4. 5. 2 Mixed and Boundary Lubrication

A mixed lubrication occurs when the ratio of the elastohydrodynamic oil film
thickness to the joint effective surface roughness is relatively small. To the hydrody-
namic friction is then added metal-to-metal solid friction at the peaks of the asperi-
ties. Both hydrodynamic and boundary conditions coexist. The surface texture con-
trols this transition from hydrodynamic to mixed lubrication, and load or speed varia-
tions or mechanical vibration may cause this transition to occur [2].

Boundary lubricated friction is calculated from [28]

F, = f0 F (4.26)

n

where f0 is the friction coefficient determined from the Stribeck diagram, and Fu is the

normal load applied on the piston ring and the cylinder bore.

55
4. 6 Results and Discussion

The instantaneous piston locations of the Diesel engine used in this study is
shown in Figure 4.4. Figures 4.5 and 4.6 illustrate the piston velocities and accelera-
tions, respectively, at each engine speed. IMEP has been calculated in order to obtain
the same power output under different engine speeds. Through the use of the P-V dia-
grams shown in Figures 4.7, 4.8, 4.9, and 4.10; computer simulations have been car-
ried out to generate the combustion gas pressures that give the same power output
used in the experimental study [51]. Table 3 shows the IMEP and power output at
each engine speed calculated through this analysis.

Table 3 IMEP and Power Output under different engine operating conditions

 

 

 

 

 

 

 

 

 

 

Engine Speed Peak Pressure Nm,x IMEP Power Output
(rpm) (kPa) (rev/sec) (kPa) (kW)
1200 8500 31.4 1356 244
1500 7600 39.3 1165 262
1800 5530 47.1 972 262
2100 4200 55.0 834 263

 

 

Piston Location(m)

Piston Velocity(m/sec)

0.20

0.15

0.10

0.05

0.00

56

 

 

I L L A l 1 L A I

 

 

200 400 600
Crank angle(degree)

Figure 4. 4 Piston location of a Diesel engine

800

 

 

 

 

 

200 400 600
Crank angle(degree)

Figure 4.5 Piston velocity at each engine speed

800

Piston Acceleration(m/sec2)

57

 

 

 

 

 

6000 Y ' I i - ' I l
——1200rpm
H. ./.\ -I
F\. ......... 15“) rpm '/ \ I. ..
\ .
4000 — -\ - - - 1800mm 1’ i ,' —
x . , \ l -
"x \. -'---2100rpm / H ,l ‘
— \\ . ‘ \'\ 01/ -t
t... \‘l (f. 1.. \'\ ’I/ _
".‘i ./. h“ l- I’-
2000 — '._\t t, . \g ., _
.I
-2000*— .\.\----’I ’\\""-’-/. "‘
.. \ I/t ‘\ / .1
I' '1
-4000 ; - A
O 200 400 600 800

Crank anq!e(aegree)

Figure 4.6 Piston acceleration at each engine speed

ylimler Pressure (Kpa)

C

(lylinrler Pressure (Kpa)

 

1 0000

8000 "I

6000 _

4000

2000

 

 

 

 

A X I I A If

 

LLLII

I

 

 

1 0000

8000

6000

4000

2000

0.005 0.010
Total Cylinder Volume(m3)

Figure 4.7 P-V diagram at an engine speed of 1200 rpm

 

 

 

 

I l

 

 

m LY

0.005 0.010
Total Cylinder Volume(m3)

Figure 4.8 P-V diagram at an engine speed of 1500 rpm

(iylinrler' Pressure (Kpa)

Cylirrrler Pressure (Kpa)

59

 

 

 

 

 

 

 

10000 ‘ ' ' ' I ' *— r f r r '
8000 — —
6000 r— -—
I- .
~ 1
4000 r- -‘
1 -1
2000 _
O I L ; I r t r
0.000 0.005 0.010 0.015

Total Cylinder Volume(m3)

Figure 4.9 P-V diagram at an engine speed of 1800 rpm

 

10000 ' ' ' ' I ' ' ' . r *
I- -I
8000 — —
)— d
6000 — _
4000 _.
2000 _
0

 

 

 

 

 

0.000 0.005 0.010 0.015
Total Cylinder Volume(m3)

Figure 4.10 P-V diagram at an engine speed of 2100 rpm

60
It has been assumed that hydrodynamic lubrication begins at the nominal mini-

mum film thickness, k=5, while boundary lubrication begins at 1:1. Also, assuming
axial symmetry throughout the circumference of the ring, the minimum oil film thick-
nesses shown in Figures 4.11 through 4.14 have been obtained with the aid of the
ring dynamics analysis program. Some unexpected changes in the curve in the middle
of the stroke are presumed to be due to the ring motion in the groove. As shown in
these figures, the minimum oil film thicknesses tend to be lower just after each TDC
and BDC. These are likely to be the points where the piston ring undergoes metallic
contact with the cylinder wall. It seems that these effects are more prominent at the
lower engine speeds than at the higher engine speeds since the combustion gas pres-
sures are relatively higher at lower engine speeds, provided that the power outputs
are the same. At the beginning of intake, and also at the end of exhaust, as shown in
Figures 4.13 and 4.14, the minimum oil film thicknesses do not approach the level at
which the mixed or boundary lubrication occurs. Presumably it would be due to the rel-
atively small difference between the pressure behind the fire ring and the combustion
gas pressure at these points. In other words, if this pressure difference is not large
enough, the fue ring pushes the cylinder wall with a relatively lower load, and thus it
results in the larger oil film thickness, which leads to hydrodynamic lubrication.

The ring friction forces are illustrated in Figures 4.15 through 4.18. As already
expected from the results of the minimum oil film thicknesses, it is clear that the high-

er the engine speed, the lower the ring friction forces at the same power output.

Minimum Oil Film Thickness(m)

Minimum Oil Film Thickness(m)

 

'- I I f T T r I Y Y q
_ -4
d

: ..
_ at
p -
p d
.- "l
,- -l
_ —-1
_ c4
_ -4
p d
_ d
_ d
_ -I
~ -
_ '1
.1

:_ _
_ d
I. -<
'- -I
p '1
’- d
_ d
p d
'- d
p -I
_ —-d
_, '4
P q
'- .1
_ CI
p d
_ 1
_ -l
b CI
.- -4
_ -—1
.— fi
.— ‘1
p d
p d
p.- d
,- -r
-q

q

d

O L A; L .L 1 I L 1 l A 4

0 200 400 600 800

61

 

 

 

 

 

0 200 400 600 800
Cronk ongle(degree)

Figure 4.11 Minimum oil film thickness at an engine speed of 1200 rpm

 

 

IIIIIIYITIII]UUIVITIiIIIIIIWIYIIYTIIrTerIIII

 

TI

 

dilllllllJilJLlleLJlll11111111llllllllllllLLllll

A A I A r L J A A 1 1 1

 

 

Cronk ongle(degree)

Figure 4.12 Minimum oil film thickness at an engine speed of 1500 rpm

Minimum Oil Film Thickness(m)

Minimum Oil Film lhickness(m)

62

 

 

 

 

 

 

 

5.0300.5 - T j r ' ' <
- -1
4.02100.5 :- J
t :
E :
3.0x10‘5 :- —3
i: -4
2.0x10'5 [- —:
5 a
1.0x10'5 :— «j
E a
O 1 u m 1 1 1 1 1 1 1 1 ‘
0 200 400 600 800
Cronk ongle(degree)
Figure 4.13 Minimum oil film thickness at an engine speed of 1800 rpm
5.0x10—5 _ r V f 1' T f T T r r r r
4.0)(10—5 E-
3.0x10'5 E-
20x10"5 '5—
1.0x10'5 5— 31
~ 3.
I 1
P -1
l—
r-
0 4' 1 1 4 1 1 1 1 i
0 200 400 600 800

Cronk ongle(degree)

Figure 4.14 Minimum oil film thickness at an engine speed of 2100 rpm

63

 

   

 

 

 

 

 

 

 

 

 

1000 T Y Y I f Y j T T I r I T I
500 — ~
’2‘ _ q
‘6 * 1
8 O A
0
LL " -l
C " .1
.9 _ ..
E» l. .
L“ -500 i- e
0" l-
.E .,
CK *- .4
—1000 '— —
‘1500 1 1 1 1 1 L 1 L 1 1 1 i 1 L 1 q
0 200 400 600 800
Crank angle(degree)
Figure 4.15 Ring fi-iction force at an engine speed of 1200 rpm
1000 _ r ' ' T ' * ' r m 1 ‘ ' r 4
i- c:
500 *- ‘r
If r- -4
if * ‘
8 O V‘A—J—J "
‘2 r d
c - q
.9 _ .1
3.3 .. 1
“- —500 ~ ~
0‘ r-
.E ‘
0: - _
i- -4
- 1 OOO 1- —
~ 1
l— .1
l- -1
L- -1
‘1500 1 1 1 l 1 1 1 i 1 1 1 1 1 1
0 200 400 600 800

Crank angle(degree)

Figure 4.16 Ring friction force at an engine speed of 1500 rpm

Ring Friction Force(N)

Ring Friction Force(N)

 

 

 

 

 

 

 

 

 

 

1000 ' . r l - . 1 . . r . . 1
r 4
500 r" _1
O V —
L- d
-500 '— ‘
—1000 r- ‘1
r .1
i— d
l. -
"1500 1 0L 1 1 1 1 1_ 1 1 1 L 1 1 1 41
0 200 400 600 800
Crank angle(degree)
Figure 4.17 Ring friction force at an engine speed of 1800 rpm
1000 _ r , r 1 . . r r . 1 , . .
500 F _
O V W F
l- _
1' -1
’ -1
1 i
1' -1
~1000 — fl
” -1
r .1
—1500 1 1 1 1 1 1 1 1 1 1 1 1 1 1
0 200 400 600 800

Crank angle(degree)

Figure 4.18 Ring friction force at an engine speed of 2100 rpm

CHAPTER 5

THE APPROXIMATE MODEL OF FIRE RING WEAR

5. 1 Relationship between Friction and Wear

There is no universal relationship between friction and wear. Relationships
can be obtained only on a case-by-case basis since both friction and wear are interfa-
cial phenomena [64, 65].

A high, steady friction coefficient may not result in high wear, but a high fric-
tion coefficient is a good indication of substantial wear. Frictional changes can be indi-
cators of changes in wear mode, so even though the friction coefficient may not corre-
late in a predictive way with wear rate, it may still be an indicator of wear transitions.
On the other hand, if the pre- and post- wear transition wear rates are known for a
given tribosystem, then the total wear volume of a sliding system can be predicted by
proportioning it in accordance with the wear process transition signaled by the fric-
tional record. This is an indirect but potentially useful technique for monitoring tribo-
systems whose mild and severe wear rates are already known [65].

The presence of third bodies, transfer layers, and film are particularly relevant
to the friction and wear relationship. It has been reported that the friction force would
change over time as deposited interfacial media assume a greater role in sliding be-
havior, particularly when the interfacial wear product removal processes in the sys-

tem are inefficient [65].

65

66
5. 2 Fire Ring Wear Model

The analysis of fire ring friction has been discussed in Chapter 4 through the
use of the Reynolds equation, the load equation, and the Stribeck diagram. Here the
procedure for determining the fire ring wear from known friction forces is presented.

An instantaneous friction work applied on the fire ring can be obtained from the

ring friction force as follows:

13wf = Pres (5.1)

where Ff denotes the friction force calculated in Section 4.5 and 88 is an instanta-

neous sliding distance on the cylinder bore. While, an instantaneous adhesive volume

loss of the fire ring can be defined as [66, 67, 68]

8VW = nFnfis (5.2)

where Fn represents the normal load applied on the fire ring, and 1] is the wear coeffi-
cient which characterizes the wear behavior of the fire ring on the cylinder wall. Equa-

tion (5.2) suggests that the wear coefficient (11) has the inverse dimension of hard-

ness or the dimension of volume over energy.

The following assumptions were made to model fire ring wear:
1) The lateral motion of the piston and the ring motion in the groove are neglected,
such that the ring friction forces are assumed to be parallel to the piston motion.
2) Axial symmetry is assumed throughout the circumference of the ring, such that cyl-
inder bore distortions [69, 70] are not allowed in this model.

3) Fire ring wear is achieved by the plastic deformation of ring material.

67
4) Wear occurs only in the mixed or boundary lubrication regime [71] where the hy-

drodynarrric film breaks down.

5) In the mixed or boundary lubrication regime, the amount of volume loss that occurs

on the ring surface is proportional to the friction work applied on the ring.

In order to find a functional form for the wear coefficient, it is necessary to in-
vestigate the factors that affect ring wear. Particularly under mixed or boundary lubri-
cated conditions, wear between the ring and the cylinder bore in relative motion is de-
termined by the surface and lubricant properties, as well as the applied load and the
sliding distance. The important parameters that rrright need to be included in Equation
(5.2) are the effective nominal minimum film thickness, ring face profile, the ratio of
true contact area to the nominal contact area, oil viscosity, and the hardness of the

ring [2, 68]. Hence, it can be deduced that the wear coefficient would be a function of

these system variables.

T1 = “(£1 H, A" K! 5) (5’3)

where C = ring face profile
11 = oil viscosity
7t = hf IO’ = effective nominal minimum film thickness
= joint effective surface roughness of the two contacting surface
hf = elastohydrodynamic oil film thickness on the cylinder bore
1: = At /A0
At = true contact area

A0 = nominal area of contact

68
a = H.111
l-lt = hardness of the ring

Hc = hardness of the cylinder

The oil film thickness under condition of hydrodynamic lubrication cannot be
easily determined since it depends upon the topography of surfaces and the height of
the asperities. Thus, the elastohydrodynarnic oil film thickness that indicates the mini-
mum oil film thickness from the surface roughness parameter and the effective nonri-
nal minimum film thickness (A) have been introduced previously [72, 73].

O’Callaghan and Provert [74] reported that the true contact area between two

abutting solids is only a fraction of the nominal area of contact as shown in Figure 5.1.

 

Surface 2 (Harder than Surface 1) /
///////////////////////////////////////////////

Figure 5.1 Schematic illustration of true area of contact

69
For the wear of minimally lubricated sliding system, Lee and Ludema [73] have ob-

served that the true contact area is linearly related to the nominal minimum film thick-

ness, and the nominal minimum film thickness is proportional to the wear rate.

111(6vw/81) cc 4110.) (5.4)

ln(1c) 0c -ln(?t) (5.5)

Since (8Vw /8t) 0c 11 , Proportionality (5.4) and Proportionality (5.5) become

11 °‘ (UK) (5.6)

1: oc (Ill) (57)

Also, it has been proposed that the adhesive wear volume is inversely proportional to
the hardness of the softer material [38, 67, 75]. Hence, using Proportionality (5.6)
and Proportionality (5.7) with the hardness ratio, the wear coefficient can be ex-

pressed as

n °c 1cm (5.8)

However, Proportionality (5.8) still presents difficulties for the quantitative analysis
of complicated ring wear phenomena. Therefore, to model the fire ring wear, a linear
relationship between the amount of volume loss and the friction work applied on the

ring is assumed throughout the engine operation. That is,

8vw oc 5wf (5.9)

70

Then, Proportionality (5.9) leads to the following expression:

11 cc fO (5.10)
where f0 is the friction coefficient determined from Section 4.5.

It has been reported that during normal engine operation, the edge of the ring
face profile is quickly worn away, allowing for the generation of a small hydrodynamic
film thickness [28]. Also, in the lubricated wear test, it has been observed that joint
effective surface roughness is decreased during running-in [76]. Thus, it can be de-
duced that the wear coefficient would be decreased during running-in. Sarkar [67, 75]
deduced an expression for the running-in wear of the machine part as an exponential
function of the sliding distance, and by using the SLA technique, Schneider et al. [51]
observed. that the wear rate of the piston ring is given by the exponential function of
the time. Hence, by using Equation (5.10) with these observations and reports, we

can define mean wear coefficient for mixed or boundary lubricated condition as
n sfota-expt—BS) +71 (5.11)
and for hydrodynamic lubricated condition as
n E 0 (5.12)

where at = (no - nwy f0

7:11.,”0

71

n”: wear coefficient at steady state

Steady state means the condition of a given tribosystem wherein the average wear
coefficient, wear rate, and other specified parameters have reached and maintained a

relatively constant level.

However, in this study coefficients a, [3, and y are evaluated from previously pub-
lished information since the wear coefficients, no and 1]”, cannot be explicitly deter-

mined from Proportionality (5.3) due to the lack of information.

Therefore, Proportionality (5.2) can be written as

8V“, = o[(1-exp(—BS) + 7] F98

= [atoexp(—BS) ”11355 (5.13)

where Ff denotes the friction force calculated in Section 4.5. Rewriting Equation

(5.13) in terms of real time for mixed or boundary lubricated conditions

5VW = [a-exp(-BUpt) + y] FfUPSt (5.14)

and for hydrodynamic lubricated condition

avw = 0 (5.15)

Thus, considering the effective nominal minimum film thickness, accumulated wear of

the fire ring can be evaluated by numerically integrating Equations (5.14) and (5.15)

72
the fire ring can be evaluated by numerically integrating Equations (5.14) and (5.15)

S. 3 Results and Discussion

The decrease in the wear rate of the piston ring has been accounted for as fol-
lows: The surface under the counterformal contact creates a situation where the load
is concentrated on a parallel narrow band of highly stressed metal. Under combined
normal and tangential stresses, the softer member of the couple flows plastically, and
the junction area grows, which results in decrease in the wear rate [75, 78]. Barber
and Ludema [77] have reported that the roughness, which would be needed in the
early stage of engine operation to enhance the removal or wearing off of the cylinder
wall and ring material, would possibly contribute to an initial high wear rate. It has al-
so been reported [67, 75] that gross surface flow occurs when sliding commences
even at a moderate load causing an increase in the hardness of the interface. A work
hardened zone also forms below the surface, and there is evidence of formation
strongly adherent oxides which protect machine parts from gross distress in services.
For instance, electron diffraction has revealed that the run-in surfaces of grey cast
iron piston rings and cylinder possess an oxide layer and graphite flakes which are

oriented with their cleavage planes parallel to the sliding distance [75].

1 hr rkrlt WerSi
Some previous papers containing useful concepts and information for the analy-
sis of ring wear are reviewed and discussed.
Suh and Saka [79] have demonstrated that the adhesive wear rate is propor-

tional to the normal load applied at the junctions for the same sliding distances.

73
Wang et al. [80] presented the mathematical model for unlubricated piston

rings shown below.

dW/dt = 0.25 waymvm (5.16)

where Kw = wear coefficient (Mpa'l)
K1 = Rum/I‘m: = temperature difference factor
Pm = the mean effective pressure (Mpa)

Vm= the mean velocity (m/sec)

However, Equation (5.16) cannot be used to calculate the ring wear for the one cycle
of engine operation because mean effective pressure and mean velocity have been
used. Moreover, the wear coefficient has been considered as a constant throughout

the engine cycles.

Sarkar [67, 75] has deduced a mathematical expression for running-in wear of

the machine part as follows :

Vw = V0[1 — exp(—ns)] (5.17)

where VW = volume loss at sliding distance 5

V0 = initial volume available at the junctions

n = proportional constant which might depends on the applied load
s = sliding distance

However, Sarkar failed to consider properly the load effect on the wear.

From empirical results, Schneider et al. [51] have presented the following

74
wear rate equation for the piston ring :

(dW/dt)mp = a-exp(—bt) + c (5.18)

where a = the wear rate at 0 hours due to break-in
b = the time dependence of break-in

c = the wear rate after break-in is complete

2ErrrEi innD' in i I
Here error estimation needs to be conducted since the coefficients a, B, and 7
used in this study were not calculated but obtained from another study [51]. Consider

that the worn volume of the ring is determined from

Vw = rtDerrww (5.19)
where Dr = ring diameter
Tr = ring thickness
WW = worn width of the ring
Equation (5.14) then becomes
6le St = [a-exp(—BUpt) + y] FfUp /(1tDrTr) (5.20)

where at, [3, and y are the coefficients dependent on the ring face profile, the oil

viscosity, the effective nominal minimum film thickness, the ratio of the true contact

75
area over the nominal contact area, and the ratio of the cylinder bore hardness over

the ring hardness as mentioned in Section 5.2. In order to estimate these coefficients,

Equation (5.18) with known coefficients a, b, and c is used in this study. Thus

a = mpg“, /(FfUP) (5.21)
B = b/Up (5.22)
y = 1th,Tr /(FfUp) (5.23)

The uncertainties in the coefficients at, B, and y can then be related to the

uncertainties involved in each parameter as follows [81] :

doc = 1301/ 8a Ida + 1301/ 8Dr IdDr + 1301/ 3Tr ldTr + I 801/ GP, 1de + 1801/ 8Up IdUP (5.24)
dB = I 313/ 311 Ian + | 6915/ BUF IdUp (5.25)
dy = by ac ldc + lay/at)r lnr)r + by art hr, + I 37/31:f I111:f + I my aUp IdUp (5.26)

Evaluating each partial derivative and substituting these into Equations (5.24),

(5.25), and (5.26),

darlat = da /a + de/Ff + dUp /Up + dDr/Dr + dTr/I‘r (5.27)
dB/B = db/b + dUp /Up (5.28)
dy/y = dc /c + de /Ff + dUp /Up + dDr/Dr + dTr/Tr (5.29)

Therefore, by specifying the uncertainties in each parameters as 10%, the

uncertainties in the coefficients at, [3, and y are estimated to be

76

dot/at = $0.5 (5.30)
dB /B = :02 (5.31)
dy/ y = 10.5 (5.32)

Attempts have been made to generate the same engine operating conditions
with the same specifications of the Diesel engine used in the experimental study
[51]. Assuming a mechanical efficiency of 90%, similar power outputs compared with
experimental data were obtained from IMEP analysis. Therefore, the coefficient B can
be determined from Equation (5.22) with the mean piston speed. However, the

coefficients at, and 7 cannot be evaluated explicitly from Equations (5.21) and (5.23)
since the friction forces depend on the lubricated condition. Thus, these coefficients
need to be determined by following procedure :

1) At the steady state, the amount of ring wear for one cycle of engine operation at
each engine speed is evaluated from the experimental study [51] by assuming
quasi-static equilibrium.

2) For one cycle of engine operation at each engine speed, iterations are necessary to

obtain the convergent wear coefficient 7.

3) Similarly, the coefficient a can be determined by considering the initial break-in

period from the experimental data.

The coefficients at, B, and 7 determined from these analysis are shown in
Table 4. Ring wear rate at each engine speed is evaluated with the average
coefficients shown in this table. Figures 5.2 through 5.5 indicate that the ring wear
rates at each engine speed for one cycle of engine operation at steady state.

Accumulated ring wear in each cases are illustrated in Figure 5.6.

77

Table 4 Wear Coefficients

 

 

 

 

 

 

 

 

 

Engine Speed Power output a B 'Y
1200 rpm 242 kW 6.2'1110'16 5,2*10-7 1,311: 10-16
warn—MW 7.71110'16 6.21404 1.7*10-16
1800 rpm 262 kW 9.511110"16 3.211110"7 2.6111016
2100 rpm 263 kW 220110-16 4.91107 521-1016
Average coefficient (11315.7)1'10-16 (4.9:1.0)*107 (2.71:1.4)*1016

 

 

 

 

Figure 5.7 illustrates the average ring wear rate at each engine speed and
power output. The results show that the higher the engine speed the wear rates are
reduced at the same power output. These trends correspond to the experimental
results [51] shown in Figure 5.8. However, as shown in these figures, the theoretical
ring wear rate does not completely comply with the experimental observations.
Presumably, this is due to the fact that the theoretical combustion gas pressures
generated by the computer simulation are not exactly the same as the experimental
data, even though each power output at each engine speed is the same. Furthermore,
it was not feasible to assign the identical values of the ring geometry, the ring face
profile, the oil viscosity, temperature distributions in the cylinder bore, and the surface
roughness of the ring and the cylinder bore used in the experiment due to the lack of
information. The selection of the point of the nominal minimum film thickness, where
the mixed lubrication begins, was estimated, so this might be an another factor which
caused errors. Also, the conversion error included in the empirical equation obtained

from experimental data cannot be neglected. As shown in Table 4, all of the data for
coefficients at and y are within the error boundary except at 2100 rpm, and the data for

1500 rpm and 1800 rpm for coefficient B are out of the error boundary. However, an

78
order of one in the specific wear rate calculation is acceptable because too much

uncertainties are involved in present analysis. Although the results shown in Figure
5.7 and Table 4 are not perfect, this is the first model of piston ring wear which

includes ring dynamics.

Ring Wear Raie(m/hr)

Ring Wear Rale(m/hr)

79

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.0x10_6 1 v r i i r 1'

)— fl '1
1.5x10_6 —- ‘
1.0x10"6 -- -
5.0x10—7 1- _

O 1 1 14A 11$ 1 1 1
0 200 400 600 800
Crank angle(degree)
Figure 5.2 Ring wear rate for one cycle of engine operation at 1200 rpm
2.0x10-6 f l ’ fi' I Y Y r r
1.5x10‘6 —- _

r-D -1

_ 1
1.0x10"ES — —
5.0x10‘7 — .1

_ 1

_ 1 1

0 AA L A 1 1 1 A 1 1 L
0 200 400 600 800

Crank angle(degree)

Figure 5.3 Ring wear rate for one cycle of engine operation at 1500 rpm

Ring Wear Rate(m/hr)

Ring Wear Role(m/hr)

2.0><10-

1.5x10'

1.0x10-

5.0x10‘

2.0x10-

1.5x10'

1.0x10’

5.0x10’

80

 

 

 

 

 

 

 

I Y I Y Y I r Y Y

6 _ .1
6 _ _
7 f_ a

1. —l
O 1 1 1 L 1 1 1 1 1 n .41 1 1 1 1

0 200 400 600 800

Crank angle(degree)

Figure 5.4 Ring wear rate for one cycle of engine operation at 1800 rpm
6 . 1 , T r , f

 

 

 

6 __ _
7 __ _

r- —1
O 1 1 1 1 1 . 1 1 1 1 1 1 1 1

O 200 400 600 800

Crank angle(degree)

Figure 5.5 Ring wear rate for one cycle of engine Operation at 2100 rpm

2.0x10'

L5x10-

1.0x10'

Ring Wear(m)

5.0x10‘

12

12

12

13

81

 

— _.-.- 2100 rpm (263 10!!)

 

. . I . r 1 1 . . . I
_ 1200 rpm (244 Kw)

1500 rpm (262 Kw)
- - - 1800 rpm (262 Kw)

 

 

 

 

200 400 600 800
Crank angle(degree)

Figure 5.6 Accumulated ring wear for one cycle of engine operation

Average Ring Wear Rate(m/hr)

Average Ring Wear Role(m/hr)

2.0x10_

1.5x10_

1.01110"

5.0x10‘

5.0x10_

82

 

I

I J7T

 

 

— 1200rpm(244 Kw) ‘
-------- 1500 rpm (262 Kw) :
-— - 1800rpm(262Kw) .
—---- 2100rpm(263 Kw) -

 

 

200 400
Engine Operating Time(hour)

Figure 5.7 Theoretical average ring wear rate at each engine speed

 

 

 

I , 1
— 1200 rpm (242 KW)
........ 1500 rpm (283 Kw)
_ _ - 1800rpm (287 Kw) «
..... 21(1) rpm (288 Kw) “

 

 

 

200 400
Engine Operating Tirne(hour)

Figure 5.8 Experimental average ring wear rate at each engine speed

CHAPTER 6

SUMMARY AND CONCLUSIONS

The flow chart shown in Figure 6.1 summarizes the entire procedure for the

fire ring wear analysis, which have been discussed so far.

A piston ring wear model has been developed through the use of the ring fric-
tion analysis with the assumption of a linear relationship between the ring wear and
the friction work applied on the surface of the ring. It was also assumed that ring wear
occurs only in the mixed or boundary lubrication regime, where the hydrodynamic film
breaks down. The lubrication regimes were separated by considering the nominal min-
imum film thickness. The ring dynamics analysis program has been used to investi-
gate the gas pressure distributions, axial motion of the ring in the groove, the mini-
mum oil film thickness on the cylinder wall, and the ring friction force under different
engine operating conditions assuming axial symmetry throughout the circumference of
the ring. Gas flow analysis shows the coupling phenomena between the ring motion in
the groove and the gas pressure distributions as predicted. Ring friction analysis dem-
onstrates that the minimum oil film thicknesses at each engine speed tend to be lower
just after TDC and BDC. Probably, these are the points where the piston ring under-
goes metallic contacts with the cylinder wall. It seems that these effects are more
prominent at the lower engine speeds than at the higher engine speeds since the com-
bustion gas pressures are relatively higher at the lower engine speeds provided that
the power outputs are the same. As a result, the higher the engine speed at the same

power output, ring friction force is reduced.
83

84
The wear rate equation for the piston ring, which is empirically obtained from

the experiment, is generalized by introducing an analytic expression in terms of ring
friction work and the wear coefficient. This ring wear analysis clearly shows that the
higher the engine speed, the wear rates are decreased when compared to the lower
speeds at the same power output. This result is consistent with the experimental ob-
servations. It seems that increasing the engine speed and maintaining the power out-
put by reducing the combustion gas pressure, if it is possible, would provide a good
protection from an unexpected failure of the piston ring due to excessive friction. In-
creased engine speeds may result in other problems, however. The effects of change
in ring face profile on the ring wear was also studied. The results shows that the ring
friction force on the parabolic face ring is less than that of barrel face ring, which gen-
erally results in less wear. Presumably, because the minimum oil film thickness pro-
duced on the parabolic face ring is higher than that on barrel face ring.

For a boundary lubricated condition, Verbeek [68] calculated the wear coeffi-
cient of the piston ring, obtaining 10'17 mzN“. Also, Childs and Sabbagh [71] have

performed boundary lubricated pin-on-ring test with cast iron materials to produce
specific wear rates of 10'19 to 10‘16 mzN'l. As shown in Table 4, the average wear co-

efficient at steady state (7) drawn from this analysis has yielded close to the maxi-
mum value of these experimental wear rate. Again, considering the uncertainties in-
volved in this analysis, it seems that this result is quite reasonable, since usually the

ring wear rate of a Diesel engine is much higher than that of an SI engine.

Though present analysis is employed locally in the Diesel engine, the model
developed in this study can be applied to any piston engine, as long as the specifica-
tions of the engine such as the oil viscosity, temperature distributions in the cylinder
bore, the ring geometry, the ratio of cylinder bore hardness over ring hardness, and

the ring face profile are given as the same. This is the first model of piston ring wear

85
which includes ring dynamics, and is intended to provide a framework for a general

model which can include the factors mentioned above.

This ring wear model should be a useful research tool in developing new mate-
rial and lubrication strategies to combat the detrimental effects of piston ring wear. A
further development of this analysis would be to generalize the model and couple it
with the experimental results of the radioactive ion implantation technique which has

been developed at MSUERL or the SLA technique developed by KFK.

F

L Estimate Inter-Rig Gas Pressure at Arbitary Crank angle

86

L START D
i

r Data Input I
i
IMEP, BP Amlysis
Piston Kinematics

F

 

 

 

]

l

 

 

l

Calculate Rateof Pressure Rise

i

 

[ Flow Rate at Next Crank agle = Previous Flow Rate

I Determination of the Axfl Motion of the Ring in the Groove

 

no\ergence

yes

—"I Calculate Mass Flow Rate for each mggGap

 

L
J‘

I Calculate New Value of Inter-Ring Gas Pressure

1

 

 

\.
<9?

yes

V

[ Evaluate Gas Pressure behind the Rings I

l

 

 

V

 

[

 

no

Determine Oil Film Pressure

    
 

Convergence

il film presthhick.

]

‘1 Evaluate Oil Film Thickness from Reynolds Equation and Load Equation ]

 

 

l Oil Film Thickness at Next Ctanlt <*a;ig:e\=Previous Oil Film Thickness ]

Figure 6.1 Procedure for fire ring wear analysis

<>941c

yes

 

TfirVI-l-"u' ‘.".'.":"5'.’.‘[n ‘ ‘ ‘t

 

87
Figure 6.1 (Cont’d)

 

L Assign 1.1 and 12 for the Determination of Lubrication Regimes I

 

 

 

 

 

 

V

[ Calculate Piston Ring Friction (Stribeck Diagram) ]
l

[ Calculate Rirgfriction Work (SW, = FfSS) I
ii

[ Ring Wear Model «iv, = nFnéS) ]

 

 

[ Define Relationship between Friction and Wear (5vw a. 8w.) 1

V
J Investigate the Effective Nominal Minimum Film Thickness (U 1

 

 

 

 

 

 

 

 

2. V :L n = fogg-exM-BSHY) 1
yes
11 = 0 i l Assign Appropriate Constants (a, B, and y) ]

 

 

v I

i
L 5V", = 0 1 L Evaluate Ring Wm Rate : 5V, /5t = (a-exp(-BSF7$F}DP I

 

 

 

 

 

 

; L Determination of Ring Wear by Numerical Integration ]

1

no kandn

 

Implementation
Completed

yes

 

Investigate the Ring Wear Rates under Different Engine Operating Conditions ]

i

[ Output Plot ]

C STOP 3

 

 

CHAPTER 7
RECOMMENDATIONS

7. l The Limitation of the Model
Axial symmetry has been assumed throughout the circumference of the ring in
this analysis. However, realistically the fire ring wears more towards the top side
than the bottom side producing an asymmetrical barrel form probably comprising two
flats or tapers [82]. This experimental result suggests that it would be possible to
predict this effect through the use of a three-dimensional model. Also, fatigue, chemi-
cal reaction, and corrosive wear were not included in this ring wear model. Such com-

plicated phenomena cannot be accounted for using this simple ring wear model.

7. 2 Recommendations for Future Research
The theoretical and experimental results indicated that a mechanism for piston
ring wear is governed by a large number of interrelated factors and should be a field
for further intensive research. Advanced analytical development work is needed in the
following areas :
1) The effect of ring design variables such as ring Width and the position of the ring
groove as well as the ring face profile on the ring wear.
2) Piston ring and bore surface finish data which can provide the surface roughness
characteristics for a more reasonable estimate of the average piston ring face
wear in an engine with different cylinder liners.

3) Further realistic piston ring lubrication theory, including lubricant flow through the

88

89
ring gaps, net transport of lubricant Within the film between the rings and cylinder

wall, and the complex movement of lubricant associated with ring lift.

4) A more sophisticated three—dimensional ring wear model including fatigue, abra-
sion and corrosive wear as well as axial, twist and rotational motion of the ring,
and cylinder bore distortion.

5) Innovative and reliable experimental methods to confirm predictions from these

models such as local ring wear, pressure behind the piston ring, ring location and

motion.

APPENDICES

APPENDIX A

THE APPLICATION OF RADIOACTIVE ION IMPLANT ATION
TECHNIQUES FOR PISTON RING WEAR MEASUREMENT

A. l The Advantages of RII over SLA

SLA (Surface Layer Activation) is limited to certain materials, e.g., iron and
chromium, where proton, deuteron, or helium ion bombardment can produce an isotope
with a reasonable half-life and an identifiable radiation. Hence, large classes of mate-
rials such as plastics, rubbers, and ceramics, (which are presently being developed for
engine parts and materials with atomic number Z < 24 ) are not accessible to this
technique. In addition, it has been suggested that SLA can be more destructive to tar-
get materials. Calculations using TRIM [1] indicate that more vacancies are created
in the SLA process than in the ion implantation for a given amount of activity [83].

Implantation of the radioactive ion for wear analysis represents a significant
advancement over the presently available SLA. The principle advantage of this tech-
nique over SLA is that instead of bombarding the target With a high energy proton
beam to change the atomic nature of the material, a heavy ion cyclotron beam is used
to implant ions into the surface. The material retains its original atomic structure, and
conveniently long-lived nucleus can be implanted. This means that a Wide class of
materials including non-metallic ceramics, graphite composites, plastics and rubber
compounds, which cannot be easily studied using SLA, can be studied using radioac-
tive ion implantation [54, 84]. The half-life of typical radionuclides produced using
SLA on such materials are too short to be useful for wear measurements.

Another unique feature of this method is that only the radioactive ion is im-

90

91
planted, and nuclear reactions do not take place in the wear sample. This eliminates

the problem of unwanted background radiation. As wear is measured in situ, this tech-
nique can be used to monitor wear rates associated with many different modes of en-
gine operation without disassembly of the engine. This method is repeatable, and
transient wearing behavior can be detected with a sensitivity of wear rate measure-

ment of rim/hr.

A. 2 The Measurement Procedure of Piston Ring Wear
The wear rate of the piston ring is measured in the following way:

1) The optimized energy of the beam and a suitable set of absorbers necessary to pro-
duce a desired uniform dose-depth profile in the surface of piston ring are obtained
using the polyenergetic TRIM simulation, which is descibed in Appendix B.

2) As shown in Figure A.l, the outer surface of the piston ring is activated using con-
dition provided from the polyenergetic TRIM simulation. It is then installed in the
piston engine.

3) The initial radiation level of the piston ring is measured by using the NaI or Ge de-
tector before the break-in. Similarly, the final radiation level is also detected after
the break-in period. Subtraction of the final activity measured from the initial activ-
ity measured gives the activity removed.

4) Assuming that the wear rate is uniform throughout the circumference of the piston
ring, it can be determined from the detector counts removed - if it is measurable -
since the radioactivity is linearly proportional to its implantation depth [45], as-

suming a uniform dose-depth distribution shown in Figure A.2.

 

93
A. 3 Radioactivity

The number of radioactive nuclei present at time t can be determined as

N = Noe-M (A.l)

where N0 is the total number of radioactive nuclei at time t = O, and l is the decay

constant. The half-life of a radioactive nucleus is then defined as the time t1,2 during

which the number of nuclei reduces to one-half the original value. Using Equation

(A.1)

1 ,2 = c‘ltlfl

which gives
it = 1n 2 / t1,2 (A.2)

This decay constant, along With the type of decay and the energy of decay, character-
izes a given radioactive nucleus and can be regarded as one of its signatures [84].

The polyenergetic TRIM simulation would be able to provide an optimized en-
ergy of the beam and a suitable set of absorbers to produce a uniform dose-depth pro-
file in the surface of piston ring. However, a problem arises concerning how many par-
ticles should be implanted into the surface of piston ring to produce a desired radiation

level. To prevent a health hazard, it should not exceed certain dosages. Suppose it is

limited Within a radioactivity of a itCi with the intensity per decay (branching ratio) of

y(%), then the number of particles to be implanted is determined as follows. The dis-

integration per second for at iiCi of this radioactive ion is

94
R = ot(3.7 * 101°)(10‘6)y (dis/sec)
= (3.7 * 104) you (dis/sec) (A.3)

Since the decay constant of a given radioisotope is given by Equation (A.2), the num-

ber of particles to be implanted is determined by

N = R / I.
= (3.7 * 104ml (1n 2 / tm)
= (5.34 * 10“)*yutt1,2 (particles) (A.4)

where y, a, and t1,2 denote an intensity per decay, activity, and half-life of the given
radioactive ion, respectively. Therefore, to produce a maximum radioactivity of a ltCi
for a specific radioisotope, the maximum number of particles allowed for implantation
are (5.34 "'104)70tt1,2 particles.

In addition, the detector efficiency needs to be calibrated to estimate an
amount of radioactivity for a given detector count. Figure A.3 shows a calibrated de-

tector efficiency [85].
Consider the radioisotope implanted into the surface of a piston ring with a

depth of d um and a dosage of (1 HQ- By using Equation (A.3) with a detector effi-
ciency of e, the mean detector count is expressed as

Cm = c (3.7 * 104 yet) (counts/sec) (A.5)
Thus the time needed to reach desired counts, Cd , is given by

td = Cd / Cm
= (2.70 * 105) cd /(yae) (sec) (A.6)

Detector Efficiency

0.013

0.012

0.011

0.01

0.004

0.002'

95

 

I

Y

 

0.001

mrm'll
5*

 

r5“-

 

 

0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5

Energy (Mev)

Figure A.3 Detector Efficiency Calibration

96
A. 4 The Sensitivity of Measurement

As stated in Section A.2, an initial and final detector count are measured for
the same period of counting time. However, these measurements involve inevitable
errors. The sensitivity related with these measurements is discussed in this Section.

Suppose that the initial detector count is measured as Ci and the final detector

count is Cf for the same period of counting time after an hour of break-in; counts re-

moved is then given by

C = c. — cf (A.7)

Thus radioactivity removed becomes

am = 0:0(Ci — Cf) / Ci (A.8)

where (10 represents an initial radioactivity.

By applying the linear relationship between dosage and depth, a depth re-

moved for an hour of break-in is expressed as

drm=d0armla0

where (10 denotes an initial implantation depth in um.

Reconsidering Equation (A.7) with probable error to check the sensitivity of

the measurement

97

 

 

 

 

 

 

 

 

 

 

 

 

.2~ /

E

B

2

C—Zo C—o C C+a Ciao
Count
Figure A.4 Normal distribution
where

with a reliability of 95 % as shown in Figure A.4. The standard deviation, 0-

given by

8C.

1

20.

I

(A. 10)

I

(A.11)

is then

98

(Si = (:i1’2 (A.12)

for counts measured, Ci [86]. Therefore, Equation (A. 10) becomes

Crm = (Ci - C.) i 2(Ci1’2 + Cfm) (A.13)

.- “51"

with a reliability of 95 %. However, this wear measurement is valid only if

Ci - Cf 2 2 (Cilf2 + Cfm) (A.14)

 

for a given period of counting time. That is,

(Cilf2 + Cfmxcfm- ail/2 + 2) s o (A.15)
since Ci1/2+Cf1/2 >0
Cflfl- C,"2 + 2 s o (A.16)
This yields
Cf s (Cim— 2)2 (A.17)

Therefore, the maximum measurable value of the final detector count is found to be

(Cil/z-Z)2 for a given initial detector count, Ci.

In addition, in order to calibrate the detector efficiency at each energy level

more accurately, errors need to be considered. A detector efficiency at each energy is

defined by

t~:=Cd/Ct (A.18)

where Cd is a detector count, and Ct is a true count in 41: steradians. While the proba-

ble error in Equation (A.18) is written as

66

25(Cd /C,)
5Cd /Ct - Cd 15Ct ICE

 

Therefore

= 5a, ch — 8C,/C, ~ (A.19)

By applying Equation (A.11) and Equation (A. 12) into Equation (A. 19)
be /c = 2 ((1 /Cd)1/2 + (1 /C,)1/2) (A20)
with a reliability of 95%. Rewriting Equation (A. 18) with error involved
2 = Cd /Ct :1: 2 Cd /C,((1 ICd)1’2 +(1/C,)1/2) (A.21)
with a reliability of 95%.

Therefore, experimental studies for in situ ring wear measurement using RII
technique expect to be performed with these estimates.

APPENDIX B

THE TRIM SIMULATION

B. 1 Introduction

A Monte-Carlo computer simulation that calculates the slowing down and
scattering of energetic ions in amorphous targets has been developed for obtaining
the most realistic range and damage profiles in any complex target [1].

This Monte-Carlo method called TRIM (Transport of Ions in Matter) as ap-
plied in simulation techniques has a number of distinct advantages over present ana-
lytical formulations based on transport theory [1, 87]. It allows the explicit consider-
ation of surfaces and interfaces, the rigorous treatment of elastic scattering with any
number of different target atoms in multi-atomic targets, and finally it yields the full
distribution function rather than a few segments of such distributions. The TRIM pro-
gram provides information on reflection and transmission properties of planar targets,
as well as ion range and collisional damage characteristics. It is also easy to include
recoil cascades, which in turn yield all the information on sputtering, ballistic mixing,
and defect production.

Several ion transport procedure based on the Monte Carlo method have been
reported [88, 89]. Aside from considering crystalline or amorphous targets, their ma-
jor differences lie in their treatment of electronic energy losses or nuclear scattering.
Since energetic ions undergo many collisions in the process of slowing down, the
method used to evaluate the scattering integral is critically important in terms of its
relative computer efficiency. Therefore, the TRIM simulation has made use of a new
analytical scheme that very accurately reproduces scattering integral results for real-

istic potentials.
100

101

B. 2 Physical Assumptions used in TRIM

The TRIM program follows a large number of individual ion or particle histories
in a target. Each history begins with a given position, direction, and energy of the ion.
The particle is assumed to change direction as a result of a sequence of collisions
with the target atom and move in straight free-flight paths between collisions. The
particle’s energy is reduced after each free-flight path by the amount of electronic en-
ergy loss and then (after the collision) by the nuclear energy loss, which is the result
of transferring momentum to the target atom in the collision [87]. Each ion’s history
is terminated either when the energy drops below a pre-specified value or when the
position of the particle has moved out of the front or rear surface of the target. The tar-
get is considered amorphous with atoms at random locations, which means that any
directional properties of the crystal lattice are ignored [1]. Thus, it actually describes
implantation into amorphous materials and neglects channelling effects that may be-
come important at low-dose low-energy implantation, which a fraction of the ions
may get steered through open passage (planar or axial) along certain directions in
crystalline structures [87, 90]. Also nuclear reactions are not included. The nuclear
and electronic energy losses or stopping powers are assumed to be independent.
Therefore, particles lose energy in discrete amounts in nuclear collision and lose ener-
gy continuously from electronic interactions. In most cases for energies below 1
MeV/amu, the electronic straggling of ion was found to be of little importance for their
projected range profiles [91]. However, electronic stopping becomes more dominant
than nuclear stopping in higher energy.

The TRIM simulations are based on the binary collision model, and thus the
ion history is determined by a series of subsequent binary encounters with the target

atoms [1]. This assumption may break down at very low energies where deflections

 

‘fi. 'Ar’i‘ _. P _. '

102
may occur even at large separations from the target nucleus. In this case the ion may

interact with more than one target atom at the same time and errors may be intro-
duced by treating such collisions separately with very small free-flight paths in be-
tween [87]. However, a quantitative study of the amount of error introduced this way

has not been performed so far.

B. 3 Preliminary Simulations and Results

Though the time integral was found to be of little importance in all cases ex-
cept for the very low energy, according to the simulation conducted, it has been ob-
served that a minimum of 1000 particles are required to get statistically reasonable
and normalized data [92]. For the heavier ion or the ion having higher energy, more
particles might need to be simulated to produce a normalized distribution since the
projected range and straggling is increased with the energy of the implanted ion.

A simulation of monoenergetic 20Ne beam into silicon nitride (Si3N4) has been

performed to check the ion concentrations and radiation damages [92]. Silicon nitride

was chosen because it is considered as one of the standards for advanced ceramic
material research. Figure 8.1 represents the 2.5 MeV/amu 20Ne ion dose-depth pro-
file implanted into Si3N4, While its radiation damage is shown in Figure 8.2.

In order to produce an uniform dose-depth profile on an implanted surface, it is
necessary to superpose the monoenergetic beam degrading their intensity by a suit-
able set of absorbers. To simulate a polyenergetic ion implantation profile from a mo-
noenergetic 20Ne beam, gold foils of varying thickness were used as degraders in

front of Si3N4. In addition, in order to perform a polyenergetic simulation, it is neces-

sary to optimize the window size of an ion range profile to achieve the desired spar-
tial resolution. If the window is too small, some of the ions might not be included in

this window due to the deflection of the ion trajectory, while if the Window is too wide,

103
although the area of interest remains the same, the spatial resolution becomes rela-

tively coarse. A spatial resolution of 50 nm has been achieved using a 5 pm window.

Figure B.3 shows the 2.5 MeV/amu 20Ne ion implanted into Si:,,N4 using different

thickness of gold foil. Six separate degrader foils having a 6 to 8.5 pm thickness of
gold fails with an interval of 0.5 pm were modeled in this polyenergetic simulation.

The simulation of a nonradioactive 20Ne ion was chosen since this beam is
easily produced at the NSCL (National Superconducting Cyclotron Laboratory), and
ion ranges and damage produced by this beam are expected to be similar to that of the
22Na beam because of its proximity in atomic number [92]. Further studies of the pro-
duction mechanism will provide guidance for realistic model development regarding

the 22Na beam.

104

 

 

 

0.00045
2 0.01134 ‘
E!
d I
F 9.53935 " U
m
<
O.
[L EN P
O
M
9%? (i.eaazsr “
2
m 8.01332)
11.] 8.00015) 1’
>
i:
:3 (me1>
DJ
M

 

 

 

 

PENETRATION DEPTH (MICRON )

Figure B.1 The ion range profile of a 2.5 MeV/amu 20Ne beam implanted onto Si3N4

 

0.04

0.03 P

0.03)

0.025 ’

0.02L

VACANC Y/A

0.013 ’

 

 

 

0.005 r II

0 1 l 1 1 —4L—‘——

14 16 18
PENETRATION DEPTH (MICRON )

Figure B2 The radiation damage of Si3N4 for a 2.5 MeV/amu 20Ne beam

 

 

RELATIVE DOSE (NUMBER OF PARTICLE/A)

105

 

20'

15-

is)

 

 

O

l" . .. .. M. 1w" L

1*—

 

 

 

 

 

 

 

 

 

 

L

 

 

2 4

PENETRATION DEPTH (MICRON )

Figure B.3 Polyenergetic profile for a 2.5 MeV/amu 20Ne beam degraded by

stacked gold foils

BIBLIOGRAPHY

BIBLIOGRAPHY
1) Ziegler, J. F., Biersack, J. P., and Littmark, U., Vol.1, The Stopping Power and
Range of Ions in Solids, Pergamon Press, NewYork (1985).

2) Heywood, J. 8., Internal Combustion Engine Fundamentals, McGraw-Hill Compa-
ny, NewYork (1988).

3) Taylor, C. F., The Internal Combustion Engine in Theory and Practice, M.I.T.
Press, Cambridge (1985).

4) McCormick, H. E., Anderson, R. D., Mayhew, D. J., and Rychlewski, W. J., "New
Developments in Piston Ring for the modern Diesel Engine", SAE Paper 750769
(1975).

5) Prasse, H. P., McComrick, H. E., and Anderson R. D., ’New Piston Ring Innova-
tions to Help Control Automotive Engine Emissions", SAE paper 730006 (1973).

6) Sunden, H., and Schaub, R., "The Manufacture of Piston Rings for Slow and Medi-
um Speed Diesel Engines", Tribology International, Vol.12, No.1, pp.3-l4, 1979.

7) Taylor, B. J., Eyre, T. S., "A Review of Piston Ring and Cylinder Liner Materials",
Tribology International, Vol.12, No.2, pp.79-89 (1979).

8) Hill, 8. H., and Newman, B. A., "Piston Ring Design for Reduced Friction", SAE
Trans. 841222 (1984).

9) Holt, J. W., and Murray, E. J., "A Two Ring Piston for Gasoline Engine", SAE
Trans. 860248 (1986).

10) Young, W. B., and McComb, J. A., "New Piston Ring Face Coating Using Design
of Experiments", SAE Trans. 900588 (1990).

11) Yoshida, H., Kusama, K., and Sagawa, J., "Effects of Surface Treatment on Piston
Ring Friction Force and Wear", SAE Trans. 900589 (1990).

12) Jeng, Y. -R., "Friction and Lubrication Analysis of Piston Ring Pack", SAE Paper
920492 (1992).

13) Furuhama, S., and Takiguchi, M., Tomizawa, K., "Effect of Piston and Piston Ring

106

107

Designs on the Piston Friction Forces in Diesel Engines", SAE Trans. 810977
(1981).

14) Furuhama, S., and Sasaki, S., "New Device for the Measurement of Piston Fric-
tional Forces in Small Engines", SAE Trans. 831284 (1984).

15) Ku, Y. G., and Patterson, D. J., "Piston and Ring Friction by the Fixed Sleeve
Method", SAE Trans. 880571 (1988).

16) Uras, H. M., and Patterson, D. J ., "Effect of Some Lubricant and Engine Vari-
ables on Instantaneous Piston and Ring Assembly Friction", SAE Trans. 840178
(1984).

17) Uras, H. M., and Patterson, D. J., "Effect of Some Piston Variables on Piston and
Ring Assembly Friction", SAE Trans. 870088 (1987).

18) Slone, R., Patterson, D. J., Morrison, K. M., and Schwartz, G. B., "Wear of Piston
Rings and Liners by Laboratory Simulation", SAE Trans. 890146 (1989).

19) Simmons, R. A., Dymock, K. R., and Webster, G. D., "Fire Ring Wear Assess-
ment on a 6V92TA Detroit Diesel Engine", SAE Trans. 890420 (1989).

20) Bishop, 1. N., "Effect of Design Variables on Friction and Economy", Paper 812A
presented Jan. 1964 at SAE congress (1964).

21) Ting, L. L., and Mayer, J. E., "Piston Ring Lubrication and Cylinder Bore Wear
Analysis, Part I - Theory", ASME trans., Journal of Lubrication Technology, 73-
Lub-25,1974.

22) Ting, L. L., and Mayer, J. E., "Piston Ring Lubrication and Cylinder Bore Wear
Analysis, Part II - Theory Verification", ASME trans., Journal of Lubrication
Technology, 73-Lub-27, 1974.

23) McGeehan, J. A., "A Literature Review of the effect of Piston and Ring Friction
and Lubricating Oil Viscosity on Fuel Economy", SAE Trans. 780673 (1978).

24) Rohde, S. M., "A Mixed Friction Model for Dynamically Loaded Contacts with
Application to Piston Ring Lubrication", Surface Roughness Efi’ects in Hydrody-
namic and Mixed Lubrication, Winter Annual Meeting of ASME, pp.19-50, No-
vember, 1980.

25) Ting, L., "A Review of Present Information on Piston Ring Tribology", SAE Trans.
852355 (1985).

26) Patton, K. J., Nitschke, R. G., and Heywood, J. B., "Development and Evaluation
of a Friction Model for Spark-Ignition Engines", SAE Trans. 890836 (1989).

108

27) Yagi, S., Ishibasi, Y., and Sono, H., "Experimental Analysis of Total Engine Fric-
tion in Four Stroke S.I. Engines", SAE Trans. 900223 (1990).

28) Brombolich, L. J., "Effect of Cylinder Distortions and Piston Ring Effects on Oil
Consumption and Friction Losses in Automobile Engines", Tribology Project Re-
search Report, Argonne National Laboratory, 1989.

29) Dawson, D., Economu, P. N., Ruddy, B. L., Strachen, P. J., and Baker, A. J.
S.,"Piston Ring Lubrication-Part II : Theoretical Analysis of a Single Ring and a
Complete Ring Pack", from "Energy Conservation through Fluid Film Lubrication
Technology : Frontiers in Research and Design", Winter Annual Meeting of AS-
ME, Dec 1979, pp.23-52.

30) Dawson, D., "Non-Steady State Effect in EHL", NASA Conference Publication
2300, Tribology in the 80’s, Vol.1 (1984).

31) Stewart, R. M., and Selby, T. W., "The Relationship Between Oil Viscosity and
Engine Performance - A Literature Search", SAE Trans. 770372 (1977).

32) Brown, S. R., and Hamilton, G. M., "Negative Pressure Under a Lubricated Piston
Ring", Journal of Mechanical Engineering Science, Vol. 20, No.1, pp49-57, 1978.

33) Furuhama, S., and Hiruma, M., "Some Characteristics of Oil Consumption Mea-
sured by Hydrogen Fueled Engine", ASLE Trans, Vol.34, No.12 (1977).

34) Reipert, P., and Buchta, R., "New Design Methods for Pistons", SAE Paper
810933, 1981.

35) Boisclair, M. E., Hoult, D. P., and Wong, V. W., "Piston Ring Thermal Transient
Effects on Lubricant Temperatures in Advanced Engines", ASME Conference and
Exhibition, Jan. 1989, 89-ICE-2.

36) McGeehan, J. A., "A Survey of the Mechanical Design Factors Affecting Engine
Oil Consumption", SAE Trans. 790864 (1979).

37) Economou, P. N., Dawson, D., and Baker, A. J. S.,"Piston Ring Lubrication - Part
I : The Historical Development of Piston Ring Tribology", ASME Trans., Vol.104,
January, 1982.

38) Rabinowicz, E., Friction and Wear of Material, John Wiley and Sons, Inc., Cam-
bridge, Massachusetts (1964).

39) Ryan, T. W. 111, Bond, T. J., and Schieman, R. D., "Understanding the Mechanism
of Cylinder Bore and Ring Wear in Methanol Fueled 81 Engines", SAE Trans.
861591 (1986).

 

109

40) Boehm, G., and Harrer, J ., "Nickel Coated Pistons for Improved Durability in
Knock Control Engines", SAE Trans. 900453 (1990).

41) Gerve, A., and Schatz, G., "Application of Cyclotrons in Technical and Analytical
Studies", Proc. 7th Int. Conf. on Cyclotrons and Their Applications, Birkhauser,
Bassel, 496 (1975).

42) Ettingger, K. V., Fremlin, J. H., and Askouri, N. A., "Measurement of Wear and
Corrosion by Surface Irradiation with Charged Particle Beam", Proc. 7th Int.
Conf. on Cyclotron and Their Applications, Birkhauser, Basel, 518 (1975).

43) Fritz, S. G., and Cataldi, G. R., "In Situ Piston Ring Wear Measurements in a
Medium-Speed Diesel Engine", ST LE Lubrication Engineering, Vol.46, No.6,
365 (1989).

44) Fritz, S. G., "Determining Wear in Internal Combustion Engines Using Radioac-
tive Components", SwRI Project 03-1542 (1989).

45) Evans, R., "Radioisotope Method for Measuring Engine Wear : A Thin Layer Ac-
tivation Method for the Measurement of Cam Follower Wear and Its Component
with a Neutron Activation Method", Wear, Vol. 64, 311 (1980).

46) Blatchley, C. C., and Sioshansi, P., "Measuring Engine Wear by Surface Layer
Activation : Extension to Nanometer Increments", SAE Paper 872155 (1987).

47) Schneider, E. W., and Blossfeld, D. H., "Real-Time Measurement of Camshaft

Wear in an Automotive Engine - a Radiometric Method", SAE Trans. 902085
(1990).

48) Sioshansi, P., and Armini, A. J ., "The Application of Ion Implantation for Large
Scale Treatment of Ball Bearing Elements", Spire Corporation, presented at the
lndustry-Academia-Govemment Workshop (1984).

49) Schneider, E. W., and Blossfeld, D. H., "Method for Measurement of Piston Ring
Rotation in an Operating Engine", SAE Trans. 900224 (1990).

50) Konstantinov, I.O., Leonov, A. I., and Mikheev, V. 1., "Surface Activation for
Wear Profile Studies of Piston Rings", Wear, Vol.14l, 17 (1990).

51) Schneider, E. W., Blossfeld, D. H., and Balnaves, M. A., "Effect of Speed and

Power Output on Piston Ring Wear in a Diesel Engine", SAE Trans. 880672
(1988).

52) Sheppard, S. R., and Suh, N. P., "The Effects of Ion Implantation on Friction and
Wear of Metals", Journal of Lubrication T echnolog)’. Vol.104, 29 (1982).

110

53) Schock, H. J., Chung, Y., Rachel, T., Grumman, D. S., Schalek, R., Ronningen, R.
M., McHarris, W. C., Hsieh, Y., and Gaydos, P., "Production and Implantation of
Ions for Wear Studies and Monitoring of Machining Process", MSUERL Repon‘
91-3 (1991).

54) Mallory, M. L., Ronningen, R. M., McHarris, W. C., Sherrill, B., Dardenne, Y. X.,
and Schock, H. J., "Tribology Studies: 7Be Implantation in a Ceramic Sample by

means of a Fragmented High-Energy l4N Beam", Nuclear Instruments and Meth-
od in Physics Research, B40/41, 579 (1989).

55) McHarris, W. C., Mallory, M. L., Ronningen, R. M., Dardenne, Y. X., and Schock,

H. J., "Gamma-Ray Spectroscopy of Implanted 7Be and 22Na for Tribology Stud-
ies", MSUCL—723 (1990).

56) Namazian, M., and Heywood, J. B., " Flow in Piston-Cylinder-Ring Crevices of a
Spark-Ignition Engine : Effect on Hydrocarbon Emissions, Efficiency and Power",
SAE Trans. 820088 (1982).

57) Kuo, T., Sellnau, M. C., Theobald, M. A., and Jones, J ., "Calculation of Flow in the
Piston-Cylinder-Ring Crevices of a Homogeneous-Charge Engine and Compari-
son With Experiment", SAE Trans. 890838 (1989).

58) Burnett, P. J ., "Relationship Between Oil Consumption, Deposit Formation and
Piston Ring Motion for Single-Cylinder Diesel Engines", SAE Paper 920089
(1992).

59) Rice, W. J ., "Development of an Instrument for Real-Time Computation of Indicat-
ed Mean Effective Pressure", NASA Technical Paper 2238 (1984).

60) Furuhama, S., Hiruma, M., and Tsuzita, M., "Piston Ring Motion and its influence
on Engine Tribology", SAE Trans. 790860 (1979).

61) Furuhama, S., and Hiruma, M., "Axial Movement of Piston Rings in the Groove",
ASLE Trans., Vol.15, No.4 (1972).

62) Brombolich, L. J ., Personal Communication (1992).

63) Allen, D. G., Dudley, B. R., Middleton, B. R., and Parker D. A., "Prediction of Pis-
ton Ring-Cylinder Bore Oil Film Thickness in Two Particular Engines and Correla-

tion with Experimental Evidence", Piston Ring Scufi'ing, London, Great Britain
(1975).

64) Czichos, H., "Importance of Properties of Solids to Friction and Wear Behavior",
NASA Conference Publication 2300, Tribology in the 80’s, Vol. 1 (1984).

111

65) Blau, P. J., Friction and Wear Transitions of Materials, Noyes Publications, Park
Ridge, New Jersey (1989).

66) Wallbridge, N. C., and Dawson, D., "Distribution of Wear Rate Data and a Statis-
tical Approach to Sliding Wear Theory", Wear, Vol.119, 295 (1987).

67) Sarkar, A. D., Friction and Wear, Academic Press Inc. London LTD, London
(1980).

68) Verbeek, H. J ., "Tribological Systems and Wear Factors", Wear, Vol. 56, 81
(1979).

69) Brombolich, L. J ., and McCormick, H. E., "Tribological Aspects of Cylinder Kit
Systems in Adiabatic Engines", SAE Trans. 880018 (1988).

70) Dinaevsky, V. V., "Analysis of Distortions of Cylinders and Conformability of Pis-
ton Rings", ST LE Tribology Trans., Vol.33, No.1, 33 (1990).

71) Childs, T. H. C., and Sabbagh, F., "Boundary-Lubricated Wear of Cast Irons to
Simulate Auto motive Piston Ring Wear Rates", Wear, Vol.134, 81 (1989).

72) Johnson, K. L., Greenwood, J. A., and Poon, S. Y., " A Simple Theory of Asperity
Contact in Elastohydrodynamic Lubrication", Wear, Vol.19, 91 (1972).

73) Lee, Y. -Z., and Ludema, K. C., "The Shared-Load Wear Model in Lubricated
Sliding Scuffing Criteria and Wear Coefficients", Wear, Vol.138, 13 (1990).

74) O’Callaghan, P. W., and Probert, S. D., "Prediction and Measurement of True Ar-
eas of Contact Between Solids", Wear, Vol.120, 29 (1987).

75) Sarkar, A. D., Wear of Metals, Pergamon Press, Oxford (1976).

76) Wu, C., and Zheng, C., "Effect of Waviness and Roughness on Lubricated Wear
Related to Running-In", Wear, Vol.147, 323 (1991)

77) Barber, G. C., and Ludema, K. C., "The Break-In Stage of Cylinder-Ring Wear: A
Correlation Between Fired Engines and A Laboratory Simulator", Wear, Vol. 118,
57 (1987).

78) Tabor, D., "Friction - The Present State of Our Understanding", Journal of Lubri-
cation Technology, Vol.103, p169 (1981).

79) Suh, N. P., and Saka, N., Fundamentals of Tribology, The MIT Press, Cambridge,
Massachusetts (1978)

80) Wang, (Y., Chen, G., and Jiao, G., "Wear Prediction for Unlubricated Piston

 

112
Rings", Wear, Vol. 135, 227 (1990).

81) Somerton, C., Error Estimation Experiment, Heat Transfer Laboratory, Michigan
State University (1992).

82) Murray, E. J., "A Survey of Scuffing in Spark Ignition Engines", Piston Ring Scuff-
ing, London, Great Britain (1975).

83) Ronningen, R., Personal Communication (1991).

84) National Superconducting Cyclotron Laboratory, Research Brief - Center for
Wear Research, January (1989).

85) Rachel, T., "An Evaluation of High Energy Ion Implantation for Wear Studies",
Master’s Thesis, Michigan State University (1991).

86) Enge, H., Introduction to Nuclear Physics, Addison-Wesley Publishing Compa-
ny, Reading, Massachusetts (1966).

87) Mazzoldi, P., and Arnold, G. W., Ion Beam Modification of Insulator, Elsevier
Science Publisher B. V., Amsterdam, Netherlands (1987).

88) Robinson, M. T., and Oen, O. S., "The Channeling of Energy Atoms in Crystal
Lattices", Applied Physics Letter, Vol.2, 30 (1963).

89) Robinson, M. T., and Torrens, I. M., "Computer Simulation of Atomic-Displace-
ment Cascades in Solids in the Binary—Collision Approximation", Phy. Rev., B9,
No. 12, 5008 (1974).

90) Mayor, J. W., Erikson, L., and Davies, J. A., Ion Implantation in Semiconductors,
Academic Press Inc., London (1970).

91) Wilson, W. D., Haggmark, L. G., and Biersack, J. P., "Calculation of nuclear stop-
ping, ranges, and straggling in the low-energy region", Phy. Rev., 315, No.5,
2458 (1977).

92) Burkhardt, T., McHarris, W., Chung, Y., Lloyd, J., Rachel, T., Schock, H., Grum-
man, D., Sharek, R., Ronningen, R., and Mallory, M. L., "Developments in High
Energy Ion Implantation for Wear Analysis", MSU Status Report MSUERL-905
(1990).

   

"11111111111111ES