kmZ Á`mo{V go _mJ©Xe©Z

To Beam As A Beacon of Knowledge

TRACK MONITORING

March 2013
INDIAN RAILWAY INSTITUTE OF CIVIL ENGINEERING
PUNE 411001
 TRACK MONITORING

MARCH 2013

INDIAN RAILWAY INSTITUTE OF CIVIL ENGINEERING

PUNE 411001
 PREFACE

It is well known that the railway tracks have to be closely

monitored for their proper upkeep, maintenance as well as to
ensure safety of trains running on them. The purpose of track
monitoring is to ensure comfort to passengers as well as to
measure geometric parameters to keep these under desirable
limits of variation and standard deviation.

While oscillation monitoring systems are used to measure

lateral and vertical accelerations of a vehicle running on tracks,
the track recording cars are more specifically deployed to monitor
the track quality.

Study material on track monitoring systems is available in a

number of RDSO reports, RTR journals and UIC reports. While
RDSO reports are directly applicable to Indian Railways, other
literature is useful to open the horizons as well as for deeper
understanding of the subject.

Necessity for compiling material on this important subject was

being felt for quite some time. Shri Naresh Lalwani, Senior
Professor, IRICEN and Shri B. P. Awasthi, Chief Engineer/TMS,
Northern Railway have made sincere efforts to fulfill this
necessity by bringing out this volume. I not only hope but feel
confident that the track engineers in particular and larger fraternity
of railways in general will find it extremely useful.

Suggestions for further improvement may be forwarded to

IRICEN for incorporating in future editions.

C. P. Tayal
Director, IRICEN

(i)
 ACKNOWLEDGEMENTS

This book is the product of posting of the first author, Naresh Lalwani,
Senior Professor (Bridges-2), in IRICEN and he is indebted to the trainee
officers who gave him the opportunity to improve this subject. Contributions
from the trainee officers having rich field experience, was not only in the
form of questions and suggestions but going to the extent of educating
on the practical aspects which can never be found in books.
B. P. Awasthi, Chief Engineer, Track Management Systems, Northern
Railway, had been associated with the development of the systems of
Track recording cars, which run today and also which are to run tomorrow
– the contact type and contactless types; during his stint at RDSO. He has
been keeping in touch with the developments taking place even today
due to his interest in the subject and by virtue of development of track
management system. He is indebted to all those who have deepened
his understanding of the subject including the present team of TMM
directorate of RDSO.
It was the atmosphere provided by Director, IRICEN, which propelled us
to go beyond the stage of writing notes to writing this book.
Authors are grateful to Shri Pradeep Kumar Garg, Senior Professor (Track-
2), who read the manuscript and gave valuable suggestions and technical
insights into the subject. His experience during tenure at RDSO in the
TMM directorate was also helpful.
We would like to thank Neeraj Khare, Associate Professor for his selfless
efforts in scruitinising the menuscript and arranging the printing of the
book.
We would thank the support staff in computer section of IRICEN, Pune,
who provided the logistic support for taking out the drafts and taking care
of demands to complete this book.
Authors acknowledge the work done by RDSO in development of TRCs
and preparing the TM reports, which have been immensely helpful in
writing of this book.
Still there is every possibility errors and mistakes creeping in. It would be
thoughtful of readers to point out to these through emails or letters. We
will try to keep it organized and devise a system that these are taken care
of in the revised version.

Naresh Lalwani B. P. Awasthi

Senior Professor (Bridges-2), Chief Engineer/TMS,
IRICEN, Pune Northern Railway, New Delhi
(ii)
 TABLE OF CONTENTS

S.N. Chapter No. Name of the Chapter Page Nos.

(i) Preface (i)
(ii) Acknowledgement (ii)
(iii) Abbreviations and Definitions 1-2
1. 1 Vital Parameters of Track 3 – 12
1.1 Introduction 3-4
1.2 Parasitic Motions of Rolling Stock 4-7
1.3 Defects at a location versus 7-8
quality of track geometry of a
specified stretch
1.4 Measurement of Defects 8 - 12
Box Parasatic Movements 5
2. 2 From Manual to Mechanical 13 - 21
Measurement
2.1 Introduction 13
2.2 Manual Monitoring of Track 13 – 15
2.3 Mechanized Monitoring 15 – 17
2.4 Progression from Manual to 17 – 19
Mechanical mode of
measurement
2.5 Measurement of four parameters 19 – 21
3. 3 Development of Reporting 22 – 45
Systems
3.1 Concept of Peak based 22 – 23
measurement and SD based
measurements
3.2 Concept of the standard deviation 23 – 27
and normal distribution curve

(iii)
 3.3 Mechanical means of Track 27 - 28
Monitoring
3.4 Requirements of Ideal Track 28 - 29
monitoring Systems
3.5 Evolution of Reporting Systems 29 – 31
in Track Recording Cars
3.6 Track Defect Measurement and 32 – 45
Reporting of Results
4. 4 Understanding TRC Results 46 – 56
4.1 Results of TRC 46 - 47
4.2 Area marked A to H are for block 48 – 55
of 200m
Box RDSO reports on Track Monitoring 47 - 48
Box Importance of chord Length 55 - 56
5. 5 Oscillation Monitoring System 57 – 62
5.1 What is it and why we need this? 57
5.2 Salient Features of OMS 58
5.3 Recording of Defects 58
5.4 Classification of Track Quality 58 - 59
5.5 Frequency of Recording 59
5.6 Non-recording speed for OMS 60
5.7 Sperling’s Ride Index 60 – 61
5.8 Important points 62
Box Typical Output of OMS 62
6. 6 Maintenance Decisions Based 63 – 77
on TGI and Related Issues
6.1 Significance of Results 63
6.2 Maintenance Decision and TGI 63 – 68
6.3 Through packing of the entire 68 – 73
section of SSE (P.Way)
6.4 Running of TRC 73 – 74

(iv)
 6.5 Frequency of Track Recording 74
6.6 Oscillograph Car 74 – 76
6.7 Procedure in the event of delay 76 – 77
TRC/Oscillograph car runs
7. 7 Track Tolerances 78 – 90
7.1 Why Tolerances 78
7.2 Track Maintenance practices on 78
Indian Railways
7.3 Type of Tolerances 79 – 89
7.4 Philosophy in fixing of Track 90
Tolerances
8. 8 Offline & Printing Software and 91 – 100
TMS for Report Generation
8.1 Need for Offline printing software 91
8.2 Track Management System 92 – 100
(TRC & OMS Module)
Box Normalized SD value 100
9. 9 How TRC measures the track 101 – 113
parameters
9.1 Measurement, analysis and 101 – 102
reporting – the past
9.2 Measurement, analysis and 102 – 112
reporting – the present
9.3 Calibration of TRC 112 - 113
Box Route Data File 111
10. Annexure-I History of TRCs 114 – 118
11. Annexure-II Typical Kilometer Wise Chart - 119
Actual Result of BCT-VR section –
Speed 100 kmph time Sheet
12. Annexure-III Basic Values Of Different 120
Parameters – One Time Sheet

(v)
 ABBREVIATIONS

C&M-1 Civil and Mechanical Engineering Report – Volume 1

CFD Cumulative Frequency Diagram
CTR Composite Track Record
G Standard Deviation of gauge variations from nominal
gauge of 1676 mm for a block of 200 m
GF Ground Features
IRPWM Indian Railways Permanent Way Manual
LACC / VACC Lateral / Vertical acceleration (in terms of g – accelera-
tion due to gravity)
LOC Block No. of the 200 m block in TRC 1 Km result
MI Maintenance Instructions
OMS Oscillation Monitoring System
PSD Power Spectral Density
RDSO Research Design and Standards Organization,
Lucknow
RIV/RIL Ride Index-Vertical and Lateral
SD Standard Deviation
SPD Speed
TGI Track Geometry Index
TMM Track Machines and Monitoring Directorate of RDSO
TRC Track Recording Car
UN1L Standard Deviation of Mid-chord offset (versine) of un-
evenness of left rail on short chord for a block of 200 m
UN1R, AL1L, Defined similarly for Alignment and Twist for short chord/
AL1R, &TW1, base
UN2L Standard Deviation of Mid-chord offset (versine) of un-
evenness of left rail on long chord for a block of 200 m
UN2R, AL2L, Defined similarly for Alignment and Twist for long chord/
AL2R & TW2 base

1
 DEFINITIONS

Track monitoring – Periodic measurement of track parameters like gauge,

unevenness, alignment, cross level and twist for the purpose of monitoring
the condition of the track.
Track parameters – Relative geometry of two rails of the track in three
dimensional systems of co-ordinates.
Unevenness defect – It is defined as undulation of rail top table (in vertical
plane) measured as mid chord offset (versine) on stipulated chord or
string held between two points located on rail top table at specified points
say 3.6 m or 9.6 m or such given distance apart. It is separately measured
for left as well as right rail.
Alignment defect – It is defined as misalignment of rail gauge face (at
13-14 mm below the rail head) measured as mid chord offset (versine)
on stipulated chord or string held between two points on gauge face at
specified points say 7.2 m or 9.6 m or such given distance apart. It is
separately measured for left as well as right rail.
Gauge defect – It is defined as deviation from the nominal distance
between the inside gauge (running) faces of two rails of track at gauge
measuring point (13-14 mm below the rail head) from the design gauge
say 1676 mm for broad gauge,1000 mm for meter gauge or any such
given distance for any other gauge.
Cross level – The difference in level between the two rails of track at right
angles to the centerline of track.
Twist defect – It is defined as out of planeness of one of the four points
from the plane passing through the other three points; of an imaginary
rectangle placed on the two rails of the track. It can also be defined as
variation over a given distance of a difference in cross level of track.
Track Geometry Index(TGI) – It is a value computed by comparing the
measured standard deviation (SD) of four parameters viz. unevenness,
alignment, gauge and twist, respectively, with the SD values for very good
track and that for track needing immediate maintenance. Higher values
indicate better track.
Engine Inspection – Inspection done by inspecting official by riding in the
engine/locomotive; popularly also called Foot-Plate Inspection.
Last Vehicle (LV) Inspection - Inspection done by inspecting official by
riding in the last or rearmost vehicle of the train; popularly also called
Brake-van inspection.

2
 Chapter 1

VITAL PARAMETERS OF TRACK

1.1 Introduction
Railway system is a network of different systems like track system,
rolling stock systems, signaling systems, operational systems etc.
Proper and coherent functioning of all the systems makes smooth
running of trains possible. Track system comprises of permanent
way system, bridge system, tunnel system, formation and cutting
systems, track machine systems etc. Rail is the road on which
wheels of trains run and therefore in Western countries they call it
Railroad. As far as physical running of trains is concerned,
interaction between the rail and the wheel is what matters.
Ideally wheels would always like the rails to be in the desired profile.
And so also the rails will like the wheels not to get much disturbed
by defects in the rail profile. Marriage which though appears to be
between two individuals is not as simple due to presence of relations
on the husband’s side and the wife’s side. Similarly the relation
between the rail and the wheel is complicated due to presence of
structure on which the rail rests and the structure of the rolling stock
above the wheels. But the measurement of the health of the relation
is done by observing the behavior between the husband & wife and
similarly of track system between the rail & wheel.
As track engineers it is our job to observe the condition and behaviour
of the track by measuring the surface of rails. Rail rests on different
components like sub-soil, formation, ballast, bridges, sleepers,
fittings, rails etc. and behaviour of these components does have a
definite effect on the behaviour of rail; like the health of human being
which depends on the functioning of heart, lungs, liver, kidney etc.
As for human body we monitor the vital parameters like temperature,
pulse rate and blood pressure to judge the health; and based on
the observation of vital parameters we take corrective actions.In the
same way for the track (pair of rails) we have defined some vital
parameters to judge its health. Track or a pair of running rail also
includes those laid on Level Crossings, Points and Crossings,
Switch Expansion Joints, Bridges etc.
If the profile of the rail is uniform without undulations, the wheels

3
 ride smoothly over it. If, on the other hand, track is non-uniform then
the rolling stock undergoes forced movements in different directions.
These movements may be gradual or sudden. Sudden movements
are felt like jerks or series of jerks. These movements impinge
upon the comfort of the passengers. If these are not attended to for
prolonged periods, and track deterioration continues, these may
even result in unsafe conditions. In a way we can see these
movements / jerks as energy imparted to the rolling stock by the
track defects and vice versa and it is by and large a disruptive and
undesirable energy.
Profiles of the track in three cardinal directions define the quality of
track. So the measurements of defects in the longitudinal, transverse
and vertical directions are the vital parameters we were talking about
in the third paragraph above. These are also called the geometric
parameters of track.
1.2 Parasitic Motions of Rolling Stock
Variations in the track geometry or defects or undulations as we call
them; are bound to be there in track though the magnitude of the
defects may vary depending on the state of maintenance. Defects in
track geometry impart energy to the rolling stock which results into
its movements in different directions. These disruptive movements
to the rolling stock are also called parasitic movements. The
magnitude of energy of the movement of rolling stock in a given
direction will be a function of the magnitude of these undulations
also called track defects on one hand and the location or distribution
pattern of track defects on the other, apart from the speed and
characteristic of the rolling stock (like spring characteristics and
damping arrangements).
We can resolve the movements of the rolling stock in three
dimensional systems – X, Y and Z. The X-axis is assumed along
the direction of the movement. Y-axis is considered perpendicular
to the direction of the movement in the horizontal plane which also
contains the X-axis. The Z-axis is in vertical direction pointing towards
the sky. All the axes are assumed to be passing through the CG of
the rolling stock. Figure 1.1 shows the different axes with respect to
a typical rolling stock.

4
 PARASATIC MOVEMENTS

Figure : 1.1

If one visualizes a coach or a wagon as shown in the sketch above,

there are 3 translational and three rotational movements in the 3
cardinal directions – x, y and z. The direction x is aligned parallel to
track, y is perpendicular to track in a horizontal plane and z is vertical
pointing towards the sky and centre of earth; all three passing through
the Center of Gravity of the coach/wagon.
SN MOVEMENT NOMEN- CAUSED BY
CLATURE
a) Translational along x axis Shuttling Loose shunting
b) Translational along y axis Lurch Alignment defect
including gauge
variations
c) Translational along z axis Bouncing Unevenness of rail top
d) Rotational about x axis Rolling Cross level/unevenness
e) Rotational about y axis Pitching Unevenness
f) Rotational about z axis Nosing Alignment defect
These are called parasitic movements because they are generated in
the rolling stock by outside factors, predominantly due to track defects.
The three kinds of movements which are important are lurch, bouncing
and rolling which have been marked in bold and underlined in the
above table.

5
Having defined the axes of reference we can study the movements
or vibrations in these directions. This would give us 3 translational
movements along the three axes and 3 rotational movements about
each of the three axes. This is shown in the table in the box – Figure
1.1.
Two translational movements viz. bouncing (along Z-axis) and
lurching (along Y-axis) and one rotational movement viz. rolling (about
X-axis) are the three movements which are important for trackmen
as these are due to track defects, which matter. These movements
of rolling stock are also called parasitic movements as the energy
for vibration of rolling stock comes from undulations or defects as a
result of forward movement of train. Without any defects with the
forward movement of train there would be no such movements/
vibrations. So these movements like a parasite depend on
something outside themselves.
Bouncing is due to the unevenness of the top table or profile of the
rails, lurching is due to alignment defects either due to bad alignment
or varying gauge; and finally rolling is due to level difference between
the left rail and right rail under the four wheels of the bogie. This is
also called twist; being the difference in cross level at two points
given length apart (normally these two points should be either axle
centers apart or bogie centers apart). For this reason following four
track (vital) parameters which define the geometric health of the
track are measured.
• Gauge
• Twist
• Alignment
• Unevenness
The crux of the track monitoring lies in measuring these four
parameters as accurately as possible, analyzing the results,
presenting them in terms of certain indices and then taking corrective
action. Reporting and analyzing can be done in many ways once
raw data has been obtained.
The parasitic movement can also be exacerbated due to defects in
the rolling stock. Hence, the objective of any track monitoring system
is to find meaning in the random parasitic movements and then

6
 isolate and measure contribution of track defects to these
movements. This, in other way, means to quantify the track defects
which are contributing to these parasitic movements, which boils
down to measuring the unevenness, twist, gauge and alignment of
track.
1.3 Defects at an isolated location versus quality of track geometry
of a specified stretch
An isolated bad spot results in rolling stock getting an isolated jerk.
At some locations numbers of individual bad spots get concentrated
in small stretch of track, which we call continuous bad stretches. A
continuous bad stretch of a track results in excitation of rolling stock
causing it to vibrate at different frequencies. Both – individual bad
spots and continuous bad stretches exist in track and can cause
discomfort to passengers and can also lead to unsafe conditions, if
allowed to deteriorate. But it is the latter which can cause vibrations
of disproportionately high magnitude due to frequency of sustained
vibrations reaching near about the natural frequency of the rolling
stock. As we all have been taught in physics that if the frequency of
forced vibration of a body equals the natural frequency of vibration
resonance can occur. Resonance is a condition of vibration when
theoretically the amplitude of vibration tends to reach infinity. So
resonance is dangerous. There have been reported cases of
isolated wagons from the middle of train formation jumping off the
track and the remaining train getting reconnected and reaching the
destination without any apparent damage.These are examples of
resonance seen in the real world.
A peculiar problem of rail wheel resonance was experienced in the
Saudi Railway. They were running trains at 150 Kmph since 1985.
In 1992 drivers of General Motors make locomotives started facing
a problem of severe juddering whenever the speed of around 120
Kmph was reached. Studies, in this instant case, revealed that the
rail length, periodicity of the coil springs and sinusoidal motion of
the wheel sets tended to coincide at 110-115 Kmph. The resonance
would have died down at higher speeds but disaster would surely
have resulted before the critical speed range was overtaken. Further
investigation revealed that the UIC’s worn wheel profile had been
recently adopted as against AAR profile till then. Reverting back to
old wheel profile solved the problem. This is the phenomenon which
decides the critical wavelength of defects important for each kind of

7
 rolling stock moving at a specified speed. And this is the reason
results on two different chords are reported on Indian Railway’s
track recording cars. Brief explanation of this concept is given in the
box at the end of Chapter 4.
Individual bad spots are equally bad as they can cause isolated
jerks beyond limits and may create unsafe conditions. These can
also produce very high stresses in the track components like rails
and fastenings at those locations, leading to local failure of these
components. Identification and measurement of the magnitude of
defects at these individual bad locations is important.
But there can be stretches, which may not have bad individual defects
of appreciable magnitude but still have a series of low magnitude
defects causing the excitation of vehicle giving bad running with
high amplitude vibrations. This, however, is dependent upon the
frequency of the defects, the type of the rolling stock (its natural
frequency of vibration) and speed of the vehicle.
1.4 Measurement of defects
How do we measure the track defects? This may appear to be too
elementary for the P.Way Engineers but for those who have just
been initiated in track engineering this is necessary to understand.
Let us forget about the title of the book for a moment and try to
answer this question. How do we measure the track defects
manually? Let us talk about each of the defects.
1.4.1 Gauge – It is easiest to measure by measuring the distance between
the gauge faces of the left and right running rails at a point 13-14
mm below the rail top table (where the wheel flange touches the
gauge face of the rail). So instead of reporting the total value, we
report variation over the design (nominal) value of gauge, which is
1676 mm for Broad gauge. Thus the gauge defect is reported say
for example as +5 mm or -4 mm, which respectively means gauge
as being 5 mm slack or 4 mm tight over the nominal value of 1676
mm,the absolute values of track gauge being 1681 mm or 1672
mm respectively for those two readings.
Twist, alignment and unevenness are the defects, which are to be
measured over the base (for twist) or chord (for alignment and
unevenness). Let us understand twist first. This has been explained
in the IRICEN Book on “The Investigation of Derailments”. But for
the sake of completeness let us understand it afresh.

8
1.4.2 Twist –Take for example a table with four legs and if one leg is
different in length by 5 mm, then the table will rock. In that case one
leg of the table will never be on the same plane as the other three.
We can say that the table has tendency to twist. We can understand
this phenomenon from the sketch at figure 1.2. A table with three
equal legs 1, 2, and 4 and one leg 3 short by x mm will rock about the
diagonal A-B as shown by arrow.

Figure 1.2 Concept of Twist for table resting on rigid legs

The case of the bogie of a coach or wagon is slightly different from
the table as it has a suspension system. Let us therefore consider
the same table with leg 3 short but now fixed with springs resting on
ground as in figure 1.3 instead of the rigid legs as in figure 1.2. The
springs will get compressed due to the load on each leg. But spring
under leg 3 will be less compressed as compared to other three
legs though it would also be touching the ground. Though outwardly
it appears that all the legs are carrying equal load as all are touching
the ground, but in reality it is not so. A closer scrutiny of the springs
will reveal that spring under leg 3 is least compressed. The load
carried by spring of leg 3 will be less than the other three legs. This
is called off-loading of one leg of the table.

L+∆ x

Figure 1.3 Concept of twist for table supported on springs

9
If we go on reducing the length of leg 3 of the table situation will
come when the spring will not carry any load and after that it may be
seen hanging in the air. At that point the spring of leg 3 is called
completely off-loaded.
Same phenomenon happens with the bogie of the wagon, which
has four wheels on two axles. In case of bogie instead of the length
of the legs being different, the ground (track) may not be level. The
difference in the level of the four points on track on which four wheels
rest causes the same situation as the case of table explained above.
The springs of the bogie will have non-uniform forces in them like in
table with springs. Now, if we go on lowering one support point on
track, there will come a stage that in-spite of the weight of the wheel
acting on the spring, one wheel will just hang in air. This wheel will
then be called fully off-loaded – carrying zero weight, the weight on
the entire bogie now being taken by other three wheels in one plane.
This though is a very extreme condition but has been quoted for
better understanding of the concept.

Figure 1.4 Concept of cross level and twist

The distance between the pairs of axles is called bogie length. The
length of bogie would be different for different rolling stocks. In the
direction of movement of train, the level of track keeps on changing
along the length (of bogie). So instead of fixed twist in mm or any
other unit of length for fixed bogie length, we talk of twist in terms of
mm/m to make twist independent of bogie length. Refer figure 1.4.
To compute the twist we will introduce the concept of cross level
difference. Cross level difference is the difference in level of left rail

10
 and the right rail under the wheels of same axle viz. 1-2 or 3-4.
Suppose for axle 1-2 X-level is 5 mm left low (left rail lower than right
rail) written as 5 LL and for axle 3-4 it is 7 mm right low written as 7
RL then the twist in mm/m of track will be calculated as follows
assuming bogie length as 3.6.m.
Twist = {5mm – (-7mm)} / 3.6 m = 12 mm/3.6 m= 3.33 mm/m
This is how it is measured manually. We assume a base as bogie
length in the above example, which is taken as 3.6 m and compute
twist from the X-levels at two locations 3.6 m apart. Twist, as we
shall see later is reported in track recording results on two different
bases –one being 3.6 m and other being 4.8 m.
1.4.3 Alignment– Straightness of any surface, horizontal or vertical is
measured by measuring the deviation from a straight. We place the
straight edge or stretched string along the surface to be measured
and we measure the gap between the straight edge and the edge of
the surface to be measured at the midpoint of straight as shown in
figure 1.5.

Figure 1.5 Measurement of unevenness and

alignment on given chord
Such measurements will have to be done for left rail as well as for
right rail. Since the chord length is longer, a string is stretched
between two points on the side of rail head and gap between the
string and the rail edge at midpoint of the string is measured which
is called mid-chord offset or versine and this is the alignment defect

11
 in mm. We call this gap as versine for a given chord (length of the
string). It is reported in mm for a specified chord length. It is shown
in figure 1.7. Similarly unevenness defects too are reported as mid-
chord versines on a specified chord length.

Figure 1.6 Versine measurement (Plan of track)

Standard chord lengths on which the versines for alignment defects
are reported are 7.2 m and 9.6 m for the Track Recording Car (TRC)
on Indian Railways.
1.4.4 Unevenness – In the manner similar to alignment, we measure the
profile of the top table of rail and results are reported on chord of 3.6
m and 9.6 m. The chord instead of being placed on the side of the
rail head (gauge face normally) as in case of measurement of
alignment is placed on the rail top table. It is measured separately
for each of the two rails – left rail and right rail. This is also called
independent chord measurements as unevenness of left rail and
right rail are independent of each other. Mid-chord versines in mm
are reported as unevenness defects for a specified chord length.
We have tried to understand the concept of vital track parameters or
defects and how to measure these manually. Track monitoring by
track recording cars measures these four defects in track but faster,
doing it directly/indirectly mechanically and/or electronically and
storing the data in digital format for detailed analysis using statistical
techniques and mathematical formulae. It also does more than this
– it measures the accelerations in lateral and vertical directions –
an objective way of assessing the discomfort to the passengers
(jerks as also felt during the engine or foot - plate/last vehicle or rear
vehicle inspections). In addition two ride indices reflecting the
passenger comfort are also reported – one for lateral direction and
other for vertical direction. We will see more of this later.

12
 Chapter 2

FROM MANUAL TO MECHANICAL MEASUREMENT

2.1 Introduction
Railway track has been in existence for last 155 years and the
systems of maintenance and monitoring of its health have evolved
over this long history reshaped by the needs and the technological
advances.
The purposes of measuring of the track geometry or track defects
are
i. To assess the riding quality of track – riding comfort of the users.
ii. To identify the individual locations as well as stretches of track
requiring attention.
iii. To take decision on periodic maintenance activities.
iv. To assess the quality of maintenance inputs given.
v. To study the trend of the deterioration.
vi. To economize on the cost of maintenance
2.2 Manual Monitoring of Track
The monitoring of track in days bygone was done by man at the
speed of man by a system of regular inspections at various levels
i.e. keyman, mate, inspectors, and as a check by officers. The system
of manual track inspections which were the only means of track
monitoring in the past is also continuing even today:
1) Daily patrol by keyman in his beat. This is inviolable and it is done,
in the absence of keyman, by senior gangman nominated by mate.
It is like daily prayers in a temple/church or mosque – continuing
from the time Railways evolved in India.
2) Periodic inspection by Supervisors.
3) Periodic inspections by Assistant Engineers– Push trolley & Engine/

13
 Rear vehicle.
4) Periodic inspections by sectional Sr. DENs/DENs – Push trolley &
Engine/Rear vehicle.
5) Periodic inspections by Sr. DEN/Coordination, HQ SAG officers &
PCE.
This monitoring system has continued up to the present times even though
the track recording cars (TRC) were introduced on Indian Railways in
1960’s. The manual monitoring system was only aided by the TRC’s and
not replaced by it.
These manual inspections are no doubt necessary and indispensable
as they are for things more than track geometry. They give the feel of the
section which means a lot more than mere track geometry measurement.
These include the supervision of quality of work done by gangs, the
condition of gang tools, their grievances, the idea of condition and
deterioration of track components, the interaction with field staff, other
issues like materials lying along the track, the progress of works (even
other than track) along the line, inspection of other features of track like
curves, points & crossing, SEJ, bridges etc. and assessment of many
unquantifiable issues like the morale of the personnel, etc.
The manual inspection for defects in track served its purpose well till it
could, even with its inherent drawbacks and limitations.
However, the limitations of manual systems are:
a) It is subjective
b) It is always for track in static and floating condition (without the train
loads)
c) It is not amenable to systematic record keeping
d) It is slow and tiring
e) It is not amenable to complex analysis for the modern track structure.
a) Subjective: – means that it depends on the skill of the person and
also varies from person to person. And if one speaks of the system
it meant practically of locating high points by unaided eye of the
mate or PWM and attending the track accordingly. Good eyesight
thus became an important asset. The vehicle response to the track
is assessed by Engine and Rear vehicle inspection by noting down

14
 the jerks & undesirable movements of rolling stock. These are also
subjective as the results of these depend not only upon the
characteristics of the vehicle and its suspension, its frequency of
vibration, but also on mental state, what we commonly call the mood,
of the inspecting official.
b) Floating condition: - In manual system the track measurements are
taken in floating condition. The behavior of track in loaded condition
is different from the floating condition as the gaps may exist between
the fittings and rail & sleeper, between sleeper & ballast and as
compressible voids in the ballast. This would make the track defect
measurements in floating condition different from what the track
would be under loaded conditions.
c) The manual recording is done on paper or even on rail foot by
means of chalk and then transferred to registers, which is difficult to
analyze owing to the huge quantum of data.
d) Slow & tiring : - The speed of measurement of track parameters is
limited by the speed of man. More so it is a back breaking work
literally and for this reason the general tendency is to avoid taking
the numerous data and there is a tendency to make up the records.
e) Not amenable to complex analysis: - And with the present
requirements of decision making speed, the manual method of
track measurement and analysis is just not enough. It is not that
this change from manual to machine came about overnight but it
evolved gradually not only because of need but also because of
technical advancement in the fields of converting of mechanical
measurement to electronic signals as also of development of
microprocessors and data storage devices.
2.3 Mechanized Monitoring
With the development of technology, mechanical means of
measurement was adopted on the Indian Railways. The
characteristic features of the mechanized measurements are
a. Direct/Indirect measurement of the 4 track parameters
b. Track parameters reported on pre-determined chord/base and
analyzed
c. Track vehicle response measured in terms of accelerations and
Ride Index

15
d. Computation of different indices like CTR, TGI etc.
At this juncture it would be proper to understand one vital fact. When
we talk of measurement of the track parameters, the accelerations
and the ride index, in mechanized monitoring it is important to
remember that it is for ensuring reasonable passenger comfort.
The discussions in the entire book are from the consideration of
ensuring minimum level of passenger comfort. The limits of track
parameters required from safety point of view are much different.
Track parameters from safety point of view: meaning safety
tolerances have not been specified for track, not only on Indian
Railways but also on the world railways.
When we talk of the passenger comfort, what we are interested in is
the disruptive energy transferred from the track-vehicle system to
the human body or to goods for that matter in the form of jerks/
vibrations. This energy will be different for different vehicles (coaches/
wagons – owing to their different oscillation characteristics and even
different maintenance conditions) even for the same stretch of the
track. Thus the most appropriate method to measure the defects
should be by measuring energy given out by a stretch of track to the
specific rolling stock and this can be done by measurement of power
spectral density. Power spectral density (PSD) describes how the
power of a signal or time series is distributed over the different
frequencies. Since it is in graphical form and is not amenable to
mathematical calculations, it is not used in track monitoring systems.
If you don’t get the hang of this, do not worry as this was just to make
you aware of the term “Power Spectral Density”. This term will not be
used anywhere else in this book. So the next question in your mind
is then what do we measure? We will come to it in the next chapter.
There are three different systems/mechanical means available with
Indian Railways to assess the track parameters and behavior, which
are
a) Track recording cars (TRC)
b) Oscillation Monitoring System or portable accelerometers (OMS)
c) Oscillograph car

16
 Out of the three, only the first one i.e. track recording cars have the
ability to measure the track parameters, whereas the other two
measure the accelerations and the ride indices. Incidentally TRC
does that too.
2.4 Progression from Manual to Mechanical mode of measurement
2.4.1 Pre-TRC System
Before TRCs were introduced in early 60s, the measurements were
done manually and data collected was too voluminous and unwieldy.
It was very difficult to draw any meaningful managerial conclusions
from it. It was looked upon more as a duty to be done religiously,
then a tool for analysis of track. It was hardly of much use for anyone
above PWI to pour over the records. It was taken in the floating
condition though ideal requirement was to do it under the train loads.
Apart from the measurement of track defects, push trolley inspections
were done to monitor the track as were engine and last vehicle
inspections to judge the quality of running. Traffic was increasing
and so were the speeds. First Track Recording Car called the
Amsler Car was imported by Indian Railways in the year 1963. A
brief history of Track Recording Cars and their development to the
modern cars would make an interesting read as well as make us
understand the present day car better and is given in the
Annexure - I.
We will talk about measurement of four parameters – Gauge, twist,
alignment and unevenness by track recording cars. We will dwell
upon accelerations and ride indices subsequently. We have already
understood in chapter 1, to measure these parameters manually.
Let us understand certain terms and concepts before we talk about
measurement of these four track parameters by TRC.
2.4.2 Concept of Sampling Interval/Sampling Distance
Let us talk about an elementary exercise of measurement of the
width of a table as shown in figure 2.1. We can measure it at the
middle of the table as at A and report the width. Alternatively we can
measure it at the two edges as at B and C and take an average. We
can also take the average of the three, or if we want to be more
precise then we can measure the width at every 30 cms and take
the average. We will get more accurate value.

17
 Figure 2.1 – Concept of sampling distance
In this case 30 cm is called the sampling distance. The same
concept is used for measurement of earthwork cross-section in
embankment construction by using sampling distance of 25 m.
When we take the measurements of track manually for new track
relaying, we talk of sleeper to sleeper variation of gauge, meaning
we have to measure gauge at every sleeper. When we want to check
the quality of track generally for a long stretch we take values of
gauge and cross level every 3-4 sleepers. All these are examples of
sampling interval. Every phenomenon of the quality control has
sampling interval or sampling frequency. In case of concrete we
take one sample for every 5 cum of concrete or part thereof and so
on. All this is done to make the sample measurements to be truer
representative of the real entity to be measured.
Since the track is a linear entity, the sampling interval is in terms of
distance. Concept of sampling interval is more important when we
are doing the measurement by some mechanical means. The
machines are obedient but dumb and have to be told everything.
Sampling interval depends on the variability of the actual parameter
along the distance apart from the accuracy required. It is also limited
by the amount of data which can be stored, handled and analyzed in
real time. The sampling interval in measurement of track parameters
was decided as 40.5 cm in the earlier track recording cars, which
are no more in operation. In the current track recording cars this has
been fixed as 30 cm.

18
2.4.3 Concept of Block
We have to have a unit length for determination of quality. We can
either determine the quality of 1 km of track or kilometer can be
further divided into shorter lengths. In the earlier TRCs, the unit of
reporting the average quality of track was 1 km. In the present day
TRCs one Km of track is further divided into units of 200 m each,
which are called blocks. So a kilometer will have normally 5 blocks
but may have 6 even if the kilometer is longer than 1000 m. This may
sound confusing from mathematical point of view but for those who
are working in the field, it is understandable. Kilometer for all practical
purposes is the actual distance between the two kilometer stones,
and not 1000m as is conventionally understood. So if actual distance
between two kilometer stones is 1125 m (long kilometer) then we
will be dividing the kilometer into 6 blocks with 5 being of 200 m
each and the 6th one being of 125 m.
Thus quality of each block is reported in the results of present day
TRC. Still there are some results for which quality is still reported
only km wise and not block wise. This is so because those results
pertain to initial system of TRCs when microprocessors were not
there and going for block-wise analysis would have been a difficult
task. And since the criteria for those results were for each kilometer,
it has been retained. The example of km as unit length for reporting
of quality is applicable,say, for categorization of track into A, B, C and
D category. This A, B, C and D will be explained later. Let us
understand more about the measurements.
2.5 Measurement of Four Parameters
2.5.1 Gauge Measurement
Gauge readings can be taken with the help of physical sensors
constantly touching the rails or contact-less sensors. We will see in
detail the way TRC actually takes various readings in subsequent
chapter. The readings for gauge are picked up at every 30 cm distance
(sampling interval). So for a stretch of 200 m around 667 (200 m/0.3
m) readings are taken and analyzed to report the quality of track for
gauge.
2.5.2 Twist Measurement
The twist can be worked out once we know the cross levels at two
ends of a stipulated base. Twist is reported on base of 3.6 m as well

19
 as on a base of 4.8 m. The word reported has been used here
instead of measured. The reason for this would become clear when
we read chapter on how TRC measures the defects. But for the
present we will call it measuring the twist. X-level is measured at
two locations 3.6 apart (or 4.8 m as the case may be) and twist is
calculated by dividing the algebraic difference of cross level by the
value of base of 3.6 (or 4.8 m). There are two ways in which this X-
level can be measured but more of it later. This measurement is
repeated by shifting the base by 0.3 m longitudinally each time.
Thus we have 667 readings (or 1334 readings of X-level) for the
twist also like for gauge for each stretch of 200m. Figure 2.2 would
help in understanding this.

Figure 2.2 – Measurement of Twist at 30 cm sample interval

2.4.3 Alignment Measurement
Alignment defect is measured by measuring the mid-chord versine
on specified chord. In case of TRC we use 7.2 m and 9.6 m chord for
reporting the alignment defect. This measurement is repeated by
shifting the base by 0.3 m longitudinally each time. Thus we have
2x667 readings (or 1334 readings – 667 for left rail and 667 for right
rail) for the alignment for each stretch of 200m. Figure 2.3 exhibits
this.

20
 Figure 2.3 – Measurement of Alignment on 7.2 m chord at 30 cm
sample interval
2.4.4 Unevenness Measurement
Unevenness is measured by taking mid-chord versine on the top
table of the rail using 3.6 m and 9.6 m chords. It is measured in the
same manner as the alignment; separately for Right Rail and Left
rail involving 2x667 numbers of readings in a block of 200 m.
Thus we have 4000 readings taken for each block of 200 m for all
the four parameters taken together.

21
 Chapter 3

DEVELOPMENT OF REPORTING SYSTEMS

3.1 Concept of Peak based measurement and SD based

measurements
Let us understand the concept of the peak based reporting as distinct
from the Average or Standard Deviation based reporting. Let us take
an example of height of students of a class. If we want to know the
tallest student or the shortest student, we will measure the height of
each of the person and report the highest and the lowest value of
height. This is peak value.
But if we have to compare the heights of students in 2 different
classes we measure the heights of students of both the classes,
work out the average height of the persons in each class and
compare to the two averages. This is an example of comparing the
averages. Instead of averages if we can also compare the standard
deviation (SD), which we can call crudely a different kind of
average.This will be called SD based comparison instead of
comparison of averages. Both have its uses.
Otherwise it becomes a case of a motorist stranded in the flood
who died in an attempt to cross the dip in the roadway by working
out the average depth of water. Water depth at the center of the dip
was more than his height. This is not to demean the importance of
average but point out its limitations. Peak measurements have its
own use as would become clear when we will understand utility of
both kinds of results.
We can understand this concept in reference to the track defects. If
we talk of quality of individual locations we are dealing with peak
values. If we talk of a quality of a stretch of track we are dealing with
average or SD values. We do not use the average values of track
defects unlike the values of average heights of class because there
is one major difference between height parameter of students and
track defects. Height is necessarily a positive value only, whereas
track defect can have either positive or negative value. This is true of
all the four track parameters except twist.

22
 Look at the sample results of gauge of two stretches of track to
understand the futility of using average values to assess quality of
track parameter.
Table 3.1 – Sample to understand futility of averages
Sample 1 Sample 2
+5 +5
-5 +5
+5 +5
-5 +5
+5 +5
-5 +5
+5 +5
-5 +5
Average = 0 Average = +5

Average gauge of sample 1 is 0, whereas average gauge of sample

2 is + 5 mm. If we see the two averages, sample 1 is very good. But
if we see the pattern of the results, in sample 1 track varies from
sleeper to sleeper and will give very bad running. Whereas the second
sample has all sleepers having uniform gauge of 5 mm slack and
should give better riding as compared to sample 1.
This brings us to a very important conclusion– “In case of track
we are more concerned with the variations in track defects rather
than the average values of the defects”. Standard deviation is the
mathematical function which serves this need. Let us understand
the concept of standard deviation before proceeding further.
3.2 Concept of the standard deviation and normal distribution curve
Indians are generally shorter in height as compared to Americans
but that does not disturb us much as we compare ourselves with
Indians mostly. Same is the case with money or possessions. So it
is the variation from the mean or average that becomes important.
Average gives us the mean of values but it does not give any idea as

23
 to what proportion of population is how far away from the mean. Or
what is the degree of variability of individuals from the mean or the
distribution of the population. We can see this in an example of
weight of people in two samples of groups of 10 persons.
σ)
Table 3.2 – Sample to understand Standard Deviation (σ

S. No. Sample 1 Sample 2

Weight xi - Χ (xi-Χ )2 Weight xi-Χ (xi-Χ )2
( xi ) Kg Kg Kg2 ( xi ) Kg Kg Kg2
1 65 -2.1 4.41 65 -2.1 4.41
2 76 8.9 79.21 87 2.9 8.41
3 98 30.9 954.81 75 7.9 62.41
4 58 -9.1 82.81 68 0.9 0.81
5 61 -6.1 37.21 73 5.9 34.81
6 59 -8.1 65.61 65 -2.1 4.41
7 52 -15.1 228.01 63 -4.1 16.81
8 55 -12.1 146.41 55 -12.1 146.41
9 88 20.9 436.81 55 4.9 24.01
10 59 -8.1 65.61 65 -2.1 4.41
Total 671 2100.90 671 306.90
Average ( ) 67.1 σ1 (SD) 14.49 Kg 67.1 σ2 (SD) 5.54 Kg

In both the samples of population size 10 the average is 67.1 Kg

even though the distribution of the population about the mean is
different as can also be seen in the pictorial representation below.
So if we take the arithmetic average of two samples it gives very
limited information about the sample.

24
Figure 3.1 – Pictorial representation of two groups of 10 samples as
per table 3.2
To get more information from a sample like the distribution of the
population about the mean we require some more information and
that is the standard deviation, which is calculated as done in the
table above. The standard deviation of the sample 1 is 14.49 Kg and
that of sample 2 is 5.54 Kg which indicates that variability of the
population about the mean is more for sample 1. The normal
distribution curves of the two populations if drawn would look like as
below. In the first case the population distribution is spread farther
away from the mean and in the second case it is clustered near to
the mean.

Figure 3.2 – Normal Distribution Curve for two groups of 10 samples

as per table 3.2

25
3.2.1 Property of Normal Distribution Curve
One more important empirical property of the normal distribution
curve is that approximately 68% of the population lies within 1
standard deviation of the mean, 95% of the population lies within 2
standard deviations of the mean and 99.7% of the population lies
within 3 standard deviations of the mean. This can be seen from the
sketch below.
The more compact curve about the mean axis (like that for sample
2 above), indicates that more number of values are closure to the
mean and lesser is the variation from the mean, hence lower value
of standard deviation and vice versa. This analogy applied to track -
lesser the value of the standard deviation fewer is the number of the
values of track defects which are away from the mean. Mean value
is the average values of the samples. Thus lesser is the SD better
is the track and vice versa.

Fig. 3.3 – Property of Normal Distribution Curve

The sample calculations for track parameter as given below in table
3.3 also bring out very clearly the futility of average and effectiveness
of the standard deviation in case of track defects.

26
Table 3.3 – Sample of track gauge readings to understand Standard
σ)
Deviation (σ
SN Parameter Parameter Difference Difference (xj-Χ )2
(Set 1) (xi) (Set 2) from average from average for set 2
(xj) of set1 (xi -Χ ) of set2 (xj-Χ )
1. -5 mm +5 mm 0 +5 25
2. -5 mm -3 mm 0 -3 9
3. -5 mm -7 mm 0 -7 49
4. -5 mm +6 mm 0 +6 36
5. -5 mm 0 0 0 0
6. -5 mm +4 mm 0 +4 16
7. -5 mm -5 mm 0 -5 25
8. -5 mm +3 mm 0 +3 9
9. -5 mm +6 mm 0 +6 36
10. -5 mm -6 mm 0 -6 36
11. -5 mm -3 mm 0 -3 9
Average -5 mm 0 mm Sum=0 Sum = 0 √Σ(xj-Χ )2 /(n)
√Σ
SD=√Σ
(x) and SD=0 =15.8/11 = 1.44 mm
Now, from the above it is clear that the 2nd set of readings has got a
higher value of SD as compared to the first set where the SD value
is 0. This is the very concept which is used in computing the SD
Values for all the 4 parameters viz. Unevenness, alignment, gauge
and twist for each block of 200 meters of track length by track
recording cars.
3.3 Mechanical means of track monitoring
The different mechanical means to assess the track parameters
and behavior as listed above are explained in brief.
3.3.1 Track Recording Cars – Controlled and maintained by TMM
directorate of RDSO – TRC measures four track parameters,
accelerations and ride indices. The maximum speed of present
TRC cars is 105 Kmph due to use of contact sensor for
measurement of gauge. In the future TRCs to be commissioned,

27
 the system of measurements of gauge will be contactless by means
of use of Laser based probes. This contactless system was there
in one of new TRC but unfortunately it was destroyed in fire during
its run.
3.3.2 Oscillation Monitoring System – under the control of either zonal
headquarters or divisions – this measures acceleration in vertical
and lateral direction and calculates the ride indices for 200 m block.
In many Railways OMS is kept in the Coach or the Loco of a
commercial train and hence does not have through run at maximum
speed of the section.
3.3.3 Oscillograph Car – run by Research directorate of RDSO – It gives
the results similar to OMS system. It runs at the maximum speed of
the section (130/120Kmph) as it is run using dedicated power with
Oscillograph Car composition as a special train.
Incidentally the accelerations sensing system and ride index
reporting is same in TRC, OMS and Oscillograph Car as the basic
equipment to measure the accelerations is the same. The only
reason we run Oscillograph (only on ‘A’ routes) is that present TRC
has a speed limitation of 105 Kmph and OMS is not able to run at
the speed of the fastest train in the section. Oscillograph Car runs
as separate train with locomotive and records the accelerations at
the maximum speed of the section, which is up to 150 Kmph in
Delhi- Agra section and 130/120Kmph on other Rajdhani routes.
Once we have sufficient number of TRCs capable of running at full
speed of the section (made possible with use of laser sensors for
gauge), Oscillograph Car run can be discontinued even on ‘A’ routes
as was done on other than ‘A’ routes.
3.4 Requirements of ideal Track Monitoring System
For a system to reflect the state of maintenance of the track only,
ideal track monitoring system should give us the track geometry
defects only and ride quality should be judged separately with the
help of accelerometers. However, both of these could be on a single
system. The requirements of an ideal track monitoring system are
i. It should show the health of the track
ii. It should be independent of the speed of recording
iii. It should not include design deviations as defects – it should

28
 have the capability to understand the design features like transition
in the curves, super elevation and versines in the circular portion of
the curve etc.
iv. It should show the effect of loading also
v. It should be recorded keeping in view the purpose of each types
of results
3.5 Evolution of Reporting Systems in Track Recording Cars
In the earlier TRCs, there was no capability for storage and analysis
of the results.The results were analogue and directly printed on the
paper. In the Amsler car, pens were used to mark the defects on a
roll of paper on which the lines for tolerance limits were drawn.
Exceedences of these tolerance limits gave an idea about the quality
of the track. Sample output was like ECG with zigzag lines on either
side of mean line. Zigzag within the narrow range from the mean
indicates better track and vice versa. Sample output for one
parameter out of 4 plotted, is given in Figure 3.4

Figure 3.4 – Sample analog chart as in Amsler Car

This mean line (dotted) for gauge represents exact gauge value of
1676 mm, for unevenness it represented level surface on a given
chord, for alignment it meant straight or curved on a given chord. For
twist it meant line of zero twist. The values of all the parameters
(except twist) could be + or – depending upon whether the gauge
was slack or tight, for unevenness whether there was high point or
dip, for alignment whether it is away from center or towards the
center of track.
To assess the quality of track or for that matter any parameter it has

29
 to be compared with the benchmark. If the passing marks of a
professional course are above 35%, 2nd division is 50%, 1st Division
is 60%, Distinction is 75% then these are the bench marks to classify
the student based on his score. Similarly track tolerances have
been specified for placing of track into A, B, C and D categories. The
value of these benchmark values have been specified in the table
given below.
For example the track irregularities for unevenness are 0–6 mm for
‘A’ category , 6-10 mm for ‘ B’, 10-15 mm for ‘C’ and >15 mm for D
category. Remember that the chords/base during the Amsler days
were only short chords. So, whenever we talk of A, B, C, and D, we
are reporting the alignment on short chord (7.2 m), unevenness on
short chord (3.6m) and twist on short base (3.6 m). On long chords,
which came quite late, these came to be designated as W, X, Y and
Z. We will see this in more detail when we discuss the typical result
of 1 km of TRC.
Table 3.4 – Limits of A, B, C and D category of track
(Para 607(1) of IRPWM
Parameter Chord/ Category and the values of defects in mm
Base (m) A B C D
Unevenness 3.6 0–6 6 (exclusive) to 10 10 (exclusive) Above 15
(inclusive) (inclusive) to 15
(inclusive)
Twist 3.6 0–5 5 (exclusive) 7.5 Above 10
(inclusive) to 7.5 (exclusive) to
(inclusive) 10 (inclusive)
Gauge - 0–3 3 (exclusive) Above +6
(inclusive) to 6 (inclusive)
Alignment 7.2 0 –3 3 (exclusive) Above 5
(inclusive) to 5 (inclusive)

Note – The values of the W, X, Y and Z though not given in the manual are
given in the first page of TRC results which gives the various limiting
values. The sample of the same is included as Annexure-III.

30
In figure 3.5, the monitor view of the current TRC can be seen, in which
the analog view in graphical form in green colour (encircled in red) for
selected 8 parameters can be seen in the bottom area of the screen. The
plot of this can also be taken but is not taken as the data summary is
available in the km wise TRC results. This is how the chart of Amsler
used to look like but was plotted live on paper. Since individual defects
were plotted these are also called peaks, this system was thus known as
peak based system.

Figure 3.5 Visual Display of output of

TRC – Analog output can also be seen
In contrast to this is the concept of Standard Deviation based system
for analysis and reporting of results, which is also known as TGI
(Track Geometry Index) based system. This was introduced in the
year 1996. There have been continuous changes/modifications in
the reporting and analysis of results from the Amsler car days to
present day TRCs.
Let us go through the evolution of the analysis and reporting systems
from 1963 to present day. Fortunately or unfortunately the present
reporting includes some of the reports prepared in pre-TGI days.
One or two results have been done away with. Each kind of result –
peak based as well as SD based have their utility. The importance
and use of different kinds of reports in the TRC results will become
clear in the chapter “Maintenance Decisions based on TGI”

31
3.6 Track Defect Measurements and Reporting of Results
Four track parameters are measured and/or computed from the
TRC run. Three track parameters were measured when the track
recording cars made their debut in 1963 and the fourth parameter
viz. alignment was introduced subsequently in 1975 with introduction
of electronic TRC.
The reporting of the results of the TRC has evolved over time with
the advancement in instrumentation and computing technology.
Different systems, indices and formulas for reporting are
summarized below.
i) A, B, C, D categorization of track
ii) CTR value in addition to categorization
iii) TQI and modified CTR value (not really implemented)
iv) Track geometry Index (TGI)
v) Worst peak information
3.6.1 A, B, C categorization of track – As explained earlier, the track
recording results were earlier available in analog form only and
analysis was done manually by counting the individual peaks for
different parameters crossing the stipulated limits. This was one of
the reasons that no formula could be evolved for representing the
Track Recording results. The track was simply classified in A, B, C
and D categories for each of the 4 parameters i.e. unevenness (on
3.6m chord), twist (over 3.6 m base), alignment (on 7.2 m chord-
calculated mathematically from 3.6 m chord) and gauge; based on
the numerical value of the track defects. The limiting values for A, B,
C and D for unevenness and twist and A, B and C for gauge and
alignment are given in Table 3.4 above. It is worth noting that D
category limits have not been specified for gauge and alignment in
IRPWM. Thus there is no D category for these two parameters i.e.
lowest category for gauge and alignment is C category. (Though
TRC result mention D category peaks for alignment and gauge too).
If you could ponder for a while, can you count the number of
parameters on which track is categorized in ABCD categories. These
are not four but 6 viz. Gauge, Twist, Alignment – Left Rail, Alignment
– Right Rail, Unevenness – Left Rail and Unevenness – Right Rail.
Each Kilometer of track was categorized in respect of each of these
six parameters.

32
 The unit of track for categorization was 1 km. It is however not that for
A Category track, for example say gauge, all the peaks should be
within the A limit of +/- 3 mm. Some tolerance, as is natural, is
allowed. A typical kilometer of track with more than 10 peaks
exceeding the limits of a particular category is classified in next
lower category for a given track parameter. Suppose at 11 locations
(we call it peaks) in a particular kilometer of track, gauge is more
than 3 mm (limit for A Category), then the track will be categorized as
B category. Up to 10 peaks exceeding a category is allowed without
downgrading the category of that kilometer of track. The track is
denoted as by a letter in prefix and a number as suffix like B6, C17 &
A9. The prefix is the category for track for that parameter as explained
above. Suffix is a number, which in all the cases, irrespective of A/B/
C/D category; indicates the number of peaks crossing the B category
limit for that particular parameter. The logic one can think of this
system for the number suffix is that ‘B’ category track is acceptable
kind of track and any comparisons with this will give an idea as to
how good or worse the track is as compared to B-category.
Table 3.5 –Practice examples for the A, B, C and D
categorization of track
BAND/Limits UN1L UNIR TW1 G AL1L AL1R
>A 31 15 63 23 39 57
>B 4 2 34 5 10 17
>C 0 0 13 2 0 2
>D 0 0 0 1 0 1
Category B4 B2 D34 B5 B10 C17

BAND/Limits UN1L UNIR TW1 G AL1L AL1R

>A 10 15 23 23 39 60
>B 4 11 12 17 25 40
>C 0 1 7 12 12 20
>D 0 0 0 1 6 15
Category A4 C11 C12 C17 C25 C40

33
Track is not designated beyond ‘C’ category for Gauge and
Alignment as per the convention and thus the worst categorization
for gauge and alignment can be ‘C’ only.
The answers of the above two practice examples are given in the
annexure III.

As already explained above, the main categorization depended upon

the peak exceedences for a particular parameter i.e. if a kilometer of
track had more than 10 peaks say for unevenness (31 peaks for first
example above) exceeding A category limit (say 6 mm for unevenness
on 3.6 m chord) but only 4 peaks exceeding B category limit, the
track would be categorized as B category for unevenness and written
as B4 (see Annexure-II). Taking example 1 of table 3.5, though the
number of peaks of Twist exceeding C category are 13 but the track
will be designated as D34 as suffix is always the number of peaks
exceeding B category irrespective of the category of the track.
Readers are requested to do the exercise for practice. As mentioned
above, the worst track categorization would be C category for
alignment and gauge unlike for twist and unevenness for which it
can be D category too.
The problem with this kind of categorization of track is that one can’t
make out how good or bad a particular Km is as compared to another
i.e. comparisons are not possible among different tracks, sub-
sections, sections etc. The reason for this is that we are trying to
make an apple-orange comparison – comparison between
dissimilar parameters. It is like you do not work out the overall score
of a student in board exams and try making comparison based on
grades (not even marks) in different subjects. It becomes a
confusing & annoying exercise. In this case if one km of track is A
category in alignment and B in twist, other km being B in alignment
and A in Twist; all other things being equal, we can’t say which is
better for two reasons. First is that we do not work out the relative
effect of alignment/twist on the passenger comfort and secondly the
magnitude of defects is not accounted for (we only count the number
once they have exceeded the threshold limits of a given category) in
the category system.
Thus, one major limitation of this system was that no meaningful
comparison between two sections of track or even two runs on
same track was possible due to incomparable data as also absence

34
 of correlation between the data. This is important from the macro
picture, which higher officials would like to have.
There is one more limitation, which can be understood from a simple
example: Track in category A4 though prima facie appears to be
better than B4 but may have peak defects of magnitude higher than
B4. For track in category A4 we know that it has 10 or less peaks
crossing A category limits and 4 peaks crossing B limits. For B4 we
know that at least 11 peaks are beyond ‘A’ category limit and out of
these 4 beyond ‘B’ category limit. But it may be possible that the 4
peaks in case of A4 track may have values beyond even the ‘D’
category limit and those in case of B4 may be marginally outside the
boundary of ‘B’ category. This may make B4 track better than A4.
Thus we can conclude the limitations of this kind of categorization
as
(i) Only number of the peaks are counted in case of exceedence
and their magnitudes are not considered
(ii) Magnitude of all samples measured are not considered for the
results
(iii) Therefore, overall health of track on any parameter is not
reflected
(iv) Since we do not have any correlation with the ride quality –
direct correlation between the categorization and the ride comfort
cannot be made
We can say that we are not able to reflect a lot of useful information
collected in A, B, C, D categorization system into the results. Secondly
we are not able to do summarization of results or work out single
figure comparison, similar to % scored by two students in board
exam by working the average marks of different subjects.
3.6.2 Composite Track Record (CTR value) - There was a need to have
the single figure results which would give an idea of the overall
condition of track (of SE’s section or AEN’s section or the division as
a whole) and also to give complete assessment of improvement or
deterioration of the track between the two TRC runs. It would also
have enabled comparison of two sections of track. Thus was evolved
the concept of Composite Track record (CTR) value.
CTR value was calculated for each kilometer using the formula as

35
below and gave the value in numbers without any units
CTR = 75 – (ULA + URA + GA + TA)
Where ULA , URA , GA & TA were number of peaks (not the magnitude
of the peaks/defects) of unevenness left rail, unevenness right rail
on the chord of 3.6 m, gauge and twist on a base of 3.6 m crossing
A category limits. This was the formula when alignment was not
measured (in the older TRCs). After the alignment measurement
was commenced the formula was modified to include alignment
defects too.
CTR = 100 - (ULA + URA + GA + TA + ALB + ARB)
ULA = No. of peaks exceeding 6 mm in unevenness on a chord of
3.6 m for left rail per km of track.
URA = No. of peaks exceeding 6 mm in unevenness on a chord of
3.6 m for right rail per km of track
GA = No. of peaks exceeding 3 mm in gauge per km of track
TA = No. of peaks exceeding 5 mm in twist on base of 3.6.m per km
of track
ALB = No. of peaks exceeding 5 mm in alignment on 7.2 m chord for
left rail per km of track
ARB = No. of peaks exceeding 5 mm in alignment on 7.2 m chord for
right rail per km of track
(Unevenness peaks and alignment peaks of both the rails are
counted; all peaks on points and crossings are counted).
Average CTR for a section of PWI or section of AEN was worked out
by simple arithmetical average value for all the kilometers in his
jurisdiction (Ref: Rly. Bds. Letter No. 90/Track-III/TK/72 dated
18.07.90).
It was being increasingly represented by the field people that CTR
value being derived from A Category defects was very sensitive to
machine measerment errors and a small deterioration in track would
have drastic reduction (highly disproportionate to the real
deterioration) in CTR value. Considering this revision of CTR formula
to increase the weightage of B Category defects was deliberated.
Revised CTR formula was framed to improve upon the CTR formula

36
(item 840 of 66th TSC)
R C T R = 1 0 0 - ( U L B+ U R B+ T B+ G B+ 0 . 5 A L B+ 0 . 5 A R B) -
0.25(ULA+URA+TA+GA)
But by the time this was approved, instructions for TQI/TGI were
also issued making it irrelevant. The system of reporting of CTR
value itself was practically discontinued even though TRC still can
give the CTR value if the option is selected in the software.
Introduction of CTR didn’t help much except for what we may call
“Indices lip service”. This did not make any meaningful change in
the reporting of results and had its own drawbacks. This system
was not a substitute of A, B, C & D categorization and was an
additional number which was easy to comprehend and reduced the
amount of data to a single value for use of higher management but
it had its own limitations. It, however, did not help to reduce the
major drawbacks of ABCD categorization.
a) Since value of CTR was based on peaks of ‘A’ category limit on
unevenness, gauge and twist and ‘B’ category for alignment any
minor deterioration in track caused significant increase in the
number of peaks crossing these strict limits. This was though
attempted to put right by RTCR which was not implemented.
b) Any defect in the wheels or in the TRC even though minor, caused
fluctuations on the CTR value of the same track from one run to
another run. It was only for very well maintained track falling in
category ‘A’ that CTR value was reasonably consistent.
Take a hypothetical example of a track having gauge parameters
in ‘A’ limit (+3 mm). Now suppose the gauge deteriorates slightly
or there is an error in the measuring system resulting in gauge at
200 locations in a Km exceeding ‘A’ limits – a minor deterioration
of 1 mm or so beyond ‘A’ limits. The CTR value of +40 suddenly
becomes -150. The deterioration in CTR value is -300%. But
has the track really deteriorated so badly? Has the running
become so rough?
c) CTR value also did not give any proper assessment of the track
for taking maintenance decisions as no limits could be laid down
for maintenance.
d) Since CTR value did not indicate the parameter in which track

37
 was bad, A, B, C and D categorization could not be done away
with.
e) CTR values were very low and most of the times had negative
values so generated a feeling of despondency. Values as low as
- 500 were reported
f) All the parameters were given equal weights.
g) It had, like ABCD categorization, no correlation with the Ride
Index.
h) And most important of all the magnitude of the defects were
still not captured in the results.
If we try to calculate the CTR value of A category track – It can be seen
that even for the track which is having unevenness, gauge and twist
parameters in A category and alignment parameter in B category,
the CTR value could be as low as 40 considering 20 peaks for
unevenness (10 each for left rail and right rail) 10 for gauge, 10 for
twist and again 20 for alignment (left rail and right rail). This is
because only when 10 peaks cross the limiting value that it can be
classified in lower category. It is like the marks for the best student
is 40%. Which student will not feel discouraged? How would one
feel telling that CTR value of my section is -20 and then to add rather
sheepishly that still my section is best in the division; and see at the
person looking at you with puzzled expression?
3.6.3 Track Quality Index (TQI) and Track Geometry Index (TGI value)
With the development of computing systems, handling of huge
amount of data became possible, not only at the post processing
stage but also processing in real time basis. The use of magnitude
of the defects in analyzing and reporting the results was possible
now. Only question was to design a reporting system using the
magnitude as well as the number of the defects in a specified length
of the track. Now with computing advances it also became possible
to calculate and derive quality of even shorter lengths of track. It was
decided to break a kilometer further into units of 200 m and report
the quality of each 200 m length of track instead of a unit of one km
of track as in ABCD and CTR system of reporting.
As we have discussed earlier, standard deviation satisfies the
requirements of analysis. In view of this it was considered desirable

38
to switch over from peak based reporting to Standard Deviation based
reporting. One of the biggest advantages the SD based system has
over the peak based system is that the magnitude of each and every
peak is taken into consideration for arriving at the value of standard
deviation of each of the four reported parameters and it also gives
an idea about the distribution of defects based on their size.
RDSO also deliberated on the SD based system of track monitoring
and came out with report TM-1 followed by TM-11 and recommended
adoption of SD based system of track monitoring on A, B, C & D
routes of Indian Railways. All the important Meter gauge routes by
this time had already been converted to broad gauge or were in the
process of conversion so report talked only of broad gauge.
With the changeover to SD based system, the microprocessor based
TRC took one set of measurements (for parameters of gauge, twist,
alignment and unevenness – last two for left rail as well as right rail)
every 405 mm (also called sampling distance) and using those
values (3000 readings in all in a block length of 200 m for the six
parameters i.e. total 1000 for unevenness – 500 each for left and
right rail, 500 for gauge, 500 for twist and again 1000 for alignment
– 500 each for left and right rail);standard deviation value for each of
these parameters was worked out for each block of 200 m.
The latest TRCs (No.7965 and 7966) use a sampling distance of
300 mm, taking 4000 readings for each block of 200 m. Standard
deviation values in mm were reported indicating the quantum of
variation in the values for a block of 200 m. Apart from reporting the
SD values, a formula was proposed to use these values of SD to
give a value called TQI (Track Quality Index) – taking into account SD
for all the four parameters. This was supposed to reflect the overall
quality of track in a 200 m stretch. TQI was never implemented and
we went for TGI.
Let us introduce a little bit of complexity here. Earlier the values of
twist were reported on 3.6 m base, alignment on 7.2 m chord and
unevenness on 3.6 m chord – these base/chords being called short
base/chords. With the change in the system of measurement and
processing, the results of SD were also reported for long base of
4.8 m for twist and 9.6 m long chord for alignment and unevenness.
Thus 11 SD values were reported for each block of 200 m.

39
Gauge – one value of SD
Twist on 3.6 m as well as 4.8 m base – two values of SD
Alignment - left rail and right rail for short chord of 7.2 m and long
chord of 9.6 m – four values of SD
Unevenness - left rail and right rail for short chord of 3.6 m and long
chord of 9.6 m – four values of SD
The formula for the TQI for each block of 200 m was
TQI = 40 * (U + 2*T +G + A)/5
Where
U = Average of the SD values of Unevenness of left and right rails
on 9.6 m chord
T = SD value of twist on 3.6 m base
G = SD value of gauge
A = Average of SD values of Alignment of left and right rails on 7.2
m chord
There were certain problems with this formula. The foremost was
that higher the value bad was the track as the TQI value was directly
proportional to the SD values. Second was that the SD values were
taken directly in the formula without any comparison with the
benchmark values. The formula can be said to be over simplistic
and crude first attempt to use the SD values. This formula was
never used for reporting as in the meanwhile a better formula was
framed removing the above shortcomings.
RDSO in their report No.TM-1 suggested the following formula for
working out TGI (Track Geometry Index). The first step was to work
out the indices for each of the individual four parameters
(unevenness, alignment, twist and gauge) for a block of 200m using
the SD values of defects on specified chords. These indices were
called GI – Gauge Index, TWI – Twist Index, ALI – Alignment Index
and UNI – Unevenness Index.
Then these values of GI, TWI, ALI and UNI are used to calculate a
composite value called TGI for a block of 200 m, which is nothing
but the weighted average indices as explained below:

40
Track Geometry Index (TGI) =

Unevenness Index (UNI) = 100xe -

Twist Index (TWI)= 100xe -

Gauge Index (GI)= 100xe -

Alignment Index (ALI)= 100xe -

TGI is calculated for a block of 200 m length of track and the notations
in the indices are explained below.
SDUm = (SDU2L + SDU2R)/2 – Average measured value of SD for
unevenness
SDU2L = Measured value of SD of Unevenness of left rail on 9.6 m
chord
SDU2R = Measured value of SD of unevenness of right rail on 9.6 m
chord
SDTm = Measured value of SD of twist on the base of 3.6 m
SDGm = Measured value of SD of gauge
SDAm = (SDA1L + SDA1R)/2 – Average measured value of SD for
alignment
SDA1L= Measured value of SD of alignment of left rail on 7.2 m
chord
SDA1R = Measured value of alignment of right rail on 7.2 m chord
SDUur = SD prescribed for urgent maintenance of unevenness at
9.6 m chord
SDUr = SD prescribed for newly laid track for unevenness at 9.6 m
chord
SDTUr = SD prescribed for urgent maintenance for twist at base of
3.6m
SDTr = SD prescribed for newly laid track for twist at base of 3.6 m
SDGur = SD prescribed for urgent maintenances for gauge

41
 SDGr = SD prescribed for newly laid track for gauge
SDAur = SD prescribed for urgent maintenance for alignment on
7.2 m chord
SDAr = SD prescribed for newly laid track for alignment on 7.2 m
chord
Note: The values of SD limits for new, urgent maintenance (Priority-
I) and planned maintenance (Priority-II) stipulated by RDSO are given
in Table 3.6 below.

Table 3.6 – SD Values for New Track, Planned (Priority-II)

and Urgent Maintenance (Priority-I)
Parameter Chord/ New Track Up to 110 Kmph - > 110 Kmph < 130 Kmph
Base tolerances SD values (mm) SD values (mm)
(m) SD values Priority II Priority I Priority II Priority I
(mm)
Unevenness 3.6 1.2 2.5 3.3 2.3 3.0
9.6 2.5 6.5 7.4 5.1 6.2
Twist 3.6 1.75 3.8 4.2 3.4 3.8
4.8 2.5 4.2 5.0 3.8 4.5
Alignment 7.2 1.5 2.7 3.8 2.3 3.0
9.6 2.5 5.0 6.3 3.3 4.0
Gauge - 1.0 2.5 3.6 1.8 2.6

Note – 1. Chord which are bold are used to calculate the UNI, TWI,
GI and ALI.

2. Priority I values pertain to urgent maintenance limits

and priority II to planned maintenance limits

As can be seen from the above table the track has been divided in
two categories based on maximum sectional speeds – (i) with
speeds less than and equal to 110 Kmph and (ii) speeds greater
than 110 Kmph but less than & equal to 130 Kmph. SD track limits
for new track are same for both the bands but the SD values for track
requiring urgent maintenance (Priority-I values) are different for the
two. The threshold limit which is 110 Kmph now was 105 Kmph
prior 2007 when the values were slightly revised. So you may find

42
value of 105 mentioned in some TRC printouts.
Having worked out TGI for a block of 200 m, the TGI for a KM is
worked out by taking mathematical average of the TGI value of 5
blocks as below. Same is to be done for computing the TGI value for
PWI’s section, AEN’s section, DEN’s section and so on.
TGI for a Km= (Sum of TGI of all the 5 blocks of KM )
5
As you can see, the measured SD values have been compared with
the benchmark values of new track and the run down track requiring
urgent maintenance. In a simpler way we can understand the formula
for computing the index value for different parameters in this way.
UNI = 100 x e-x

We can write the formula for index value for other parameter TWI, GI
and ALI in a similar manner? The formula is written in this way for
easier understanding.
For a moment let us consider a track in such a good condition of
maintenance that the measured SD for unevenness is equal to the
SD for new. The value of the x will become 0 (zero) and e to the
power 0 would be 1 giving the value of UNI as 100. We can conclude
from this that for track maintained to new track standards for a
particular parameter, the value of index for that parameter would be
100.
Let us a case of a rundown track badly in the need of maintenance
such that its SD value has increased and is equal to SD value for
track requiring urgent maintenance. The value of the x will become
1 and e to the power -1 would be around 0.36 giving the value of UNI
as 36. We can conclude from this that for track deteriorated to
standards of track requiring urgent maintenance for a particular
parameter, the value of index for that parameter would be 36. For the
track maintained to standards in between these would have index
value between 36 and 100.
What can be the highest Index Value for any parameter? Would it be
more than 100? Would it be same for all the parameters as are the

43
 values for track maintained to new track standards i.e. 100 or that
requiring urgent maintenance i.e. 36 ? The answer to these
questions would depend on the possible value of measured SD of
the actual track. Theoretically the best possible SD value can be
zero. Taking this as measured SD, if we work out the value of indices
that will the maximum. The values will depend upon the SD for new
track as well as SD for track requiring urgent maintenance of
individual parameter and Index values for individual parameter will
be different.
The maximum value of indices and TGI is given below for each of
the speed bands for appreciation of a reader and to satisfy his
intellectual curiosity.
Table 3.7 – Theoretically maximum possible value of SD based Indices
S. No. Parameter Ideal Value Value Value Track TGI –
of meas- of SD – of SD – Indices Maxi
ured SD Urgent New Maxi mum
Mainte- Track mum–
nance Value
Speed Up to 110 Kmph
1. Gauge 0 3.6 1.0 146.90
2. Twist – 3.6 m base 0 4.2 1.75 204.27
3. Alignment – 7.2 m chord 0 3.8 1.5 191.97 184.34
4. Unevenness – 9.6 m chord 0 7.4 2.5 170.22
Speed>110 Kmph and <=130 Kmph
1. Gauge 0 2.6 1.0 186.82
2. Twist – 3.6 m base 0 3.8 1.75 234.82
3. Alignment – 7.2 m chord 0 3.0 1.5 271.83 244.57
4. Unevenness – 9.6 m chord 0 6.2 2.5 196.54

The result of TRC gives not only TGI but many other things like
ABCD categorization of track on short chords, WXYZ on long chord,
Ride Index – lateral and vertical, parameters to be monitored as per
para 607 of IRPWM, isolated defects with location and magnitude,
lateral and vertical accelerations exceeding the threshold limits. It

44
also gives indication of maintenance instruction for blocks needing
attention. It is important to fully understand the results given by TRC
for each kilometer. Normally result for one km is printed on one A4
size sheet in landscape orientation. A typical result is enclosed as
Annexure-II.
It is important to understand the results given in this one page result,
which some PWIs jokingly call the horoscope of the track.The next
chapter describes in reasonable detailabout this 1 Km TRC results.
We will also discuss in another chapter, the kind of analysis which
are possible using an offline and printing software given by RDSO.
It is slightly different for different TRCs. Now with the introduction of
web based TMS (track management system) it is possible to upload
the TRC data file in TMS at the level of the division and do all kinds
of analysis required by the field people and also higher
management. In the same chapter are discussed different kinds of
analysis possible with TMS. Though TMS might not have been
implemented in all the divisions as of now but eventually will be
done.

45
 Chapter 4

UNDERSTANDING TRC RESULTS

4.1 Results of TRC

TGI based track recording system came into vogue in TRC in the
end of 1996. The recording is based on the concept of working out
the standard deviation of track parameters by dividing a kilometer
into blocks of 200 m each. Though the system has been in existence
for last 17 years, still the understanding of the basics pertaining to
the results given is not what it should be by the field engineers.
Unfortunately not much study material is easily available for reference.
And IRPWM also does not even talk about the latest system. This
chapter deals with TRC No.7965, 7966 and 7967. Older TRC – 225
and 2500 have been phased out.
If TRC is running at a speed of 100 Kmph, then it will take 36 seconds
to cover 1 Km and 7.2 seconds to cover one block of 200 m. In this
short time it picks up 4000 readings and analyzes them by the time
TRC completes the taking of reading for next block.
Table 4.1 – Readings taken by TRC for each block of 200 m
S. No. Parameter Readings for each
block of 200 m
1. Unevenness – Left Rail 667 Nos.
2. Unevenness – Right Rail 667 Nos.
3. Twist 667 Nos.
4. Gauge
5. Alignment – Left Rail 667 Nos.
6. Alignment – Right Rail 667 Nos.
Total Readings for each block of 200 m 4000 Nos.

In addition to SD values we still need the worst peak information for

attending isolated bad points in the track. This is also given in the
TRC results. A typical result for a kilometer is given as Annexure-II,
marking the different kinds of results for understanding of each of
them.

46
 As you can see in Annexure-II, the area of the chart has been divided
into 9 areas marked A to L (there is no G to avoid confusion with G for
gauge and also no I to avoid confusion with 1). Readers are advised
to take out a photocopy of Annexure-II and III and keep it in front of
them while studying this chapter.
The readings in each of these marked areas are explained below.

RDSO reports on Track Monitoring

Track Machine and Monitoring (TMM) directorate of RDSO is the
nodal directorate for the Track recording systems. Number of
studies and trials have been conducted on different aspects of
the track monitoring starting from the legendary Civil and
Mechanical Engineering Report (Volume – I) – C&M-1 as it is
fondly known.
It was in Report No. TM-1 of Aug’94 (Evolution of Standard
Deviation (SD) Based Track Geometry Index) that the process of
change over from peak based system to SD based system was
initiated. Based on the recommendations in this report Railway
Board (RB) nominated a high level committee comprising of CE/
WR, CE/SR, CE/ER, CE/NR, ED/Track(MC)/RB as members and
Director/TMM, RDSO as convener. The committee was to go into
details of the report for implementation. This committee submitted
its recommendation in Feb’95. From the above, Report No.TM-
11 of Dec’96 was made out by RDSO for implementation of SD
based Track Indices. Railway Board issued letter No.94/Track-
III/TK/23 dated 30.12.1996 implementing the SD based track
monitoring system on Indian Railways.
Trials were separately done for fixing the SD based parameters
for speeds of 160kmph using WDM-2 Locomotive and IR-20
coach. Report No. TM-4 was published by RDSO. Not much was
further done on this.
Then in Report No.TM-109 RDSO did trials of correlating the
limiting values of the (Ride Index) RI to arrive at the permissible
urgent maintenance values (now called Priority-I values) of SD.
The values prescribed in TM-11 were revised vide RB’s letter
No.2007/Track-III/TK/7 dated 12-07-2007.

47
 Need for through packing was linked to TGI value considering
SSE’s section as a unit by drawing Cumulative Frequency
Diagram (CFD) of TGI. Values of CFD to be achieved post tamping,
to monitor the quality of tamping, were also specified in RDSO
Report No.TM-115.
There are many more reports for subjects like validation of Track
Recording Cars, trials to find out the limiting values of RI and
other parameters, track geometry standards for different speed
bands etc.
Interested readers can get these reports either from IRICEN
library/RDSO. Author can also be contacted for soft copy of some
reports, which are available.

4.2 Area marked A to H are for block of 200m.

4.2.1 Area marked A & B – The value in front of each block is the SD
values in mm for that parameter for that block except for AVG-G. The
abbreviations in the column headings are as below.
LOC – Location (Block No.),
UN1L & UN1R – Measured SD value of Unevenness defect on
short chord of 3.6 m, for left & right rail
TW1 – Measured SD value of Twist defect on short base
of 3.6 m,
G – Measured SD value of Gauge defect,
AL1L & AL1R – Measured SD value of Alignment defect on short
chord of 7.2 m, for left & right rail
AVG-G (not SD) – Measured Average value of gauge –moving
average of 50 readings
Similarly in the area marked B, the SD value in mm are given for the
long chord/base. The abbreviations in the column headings are as
below
UN2L & UN2R – Measured SD value of Unevenness defect on
long chord of 9.6 m, for left & right rail
AL2L & AL2R – Measured SD value of Alignment defect on long

48
 chord of 9.6 m, for left & right rail
TW2 – Measured SD value of Twist defect on long base
of 4.8 m
In these abbreviations, 1 stands for short chord/base and 2 stands
for long chord/base.
4.2.2 Area marked C– Indices for the four parameters Unevenness Index
(UNI), Twist Index (TWI), Gauge Index (GI) and Alignment Index (ALI)
for each block. These indices are calculated using the measured
SD values (as given in Area marked A & B), the SD value of the new
track tolerances and the SD value for the urgent maintenance (Priority-
I Values) tolerances as given in table 3.6 in the previous chapter
along with the sample calculation as below for clearer
understanding.
The Km wise chart in the annexure-II is for track section having
speed less than 110 kmph. So we calculate the UNI value for block
1 manually just to understand the process, the values of SDUm,
SDUur and SDur are as below
SDUm (from area marked B in Km wise chart) = (SDU2L + SDU2R)/
2 = (4.5 + 4.0)/2 = 4.25 mm
SDUur (from the table 3.6 above) = 7.4 mm
SDUr (from the table 3.6 above) = 2.5 mm
UNI (Unevenness Index ) = 100xe–(SDUm-SDUr)/ (SDUur-SDUr) =100xe–(4.25-2.5)/
(7.4-2.5)
= 69.96 say 70, which is the value in Annexure-II for the first
block.
Indices for other parameters can be calculated in a similar manner
for all the blocks of a kilometer.
4.2.3 Area marked D– The first column gives the TGI value of each block
and TGI value is weighted average of the four indices viz. UNI, TWI,
GI and ALI by giving weightage of 2, 1, 1 and 6 respectively. Let us do
this calculation for the 1st Block.
UNI = 70 x 2 = 140
TWI = 55 x 1 = 55
GI = 100 x 1 = 100
ALI = 42 x 6 = 252
Total 547

49
 TGI = 547/10 = 54.7 say 55 as can be seen in the area marked
D of Annexure-II for 1st block.
TGI for other blocks can be calculated similarly. TGI for a kilometer
is the arithmetic average of TGI of all the blocks in a KM.
TGI for this Km = (55 + 73 + 78 + 68 + 57) / 5 = 331/5 = 66.2 say 66
as can be seen in the encircled value seen between H and J.
The second column with heading SPD gives the average speed in
Kmph in that particular block – 33 kmph in the 1st block, 47 kmph in
the 2nd block and so on.
4.2.4 Area marked E– This gives the RIV – Ride Index Vertical value and
RIL – Ride Index Lateral value for each of the block; just the same
way we get in OMS recordings.
4.2.5 Area marked F– The heading of the column is MI – Maintenance
Instructions. There are no numerical values in this column but has
either the markings with 2 asterisks** or 3 asterisks*** against
some of the blocks or blank space.
For each block there are 6 values of standard deviation in the area
marked A (excluding the values under heading AVG-G) for the short
chords/base and 5 values of standard deviation in area marked B
for long chords/base. Even if value of one parameter out of these 11
values, exceeds the priority-II values (as given in table 3.6) of SD for
that speed band, the MI column for that block shows 2 asterisk **
mark. Similarly even if value of one parameter out of these 11 values,
exceeds the priority-I (again as given in table 3.6) values of SD for
that speed band, the MI column would shows 3 asterisk *** mark.
This is how the monitoring of SD values on long chord as well as
short chord is done.
It would be worthwhile to mention here that the short chord defects
are pertinent for goods wagon and long chord defects are more
pertinent for the coaching stock. Field engineers are supposed to
identify the worst blocks and attend these blocks and we call these
as spot attention.
Let’s see for a real result for 1st block in Annexure-II. It has a **
marking in MI column. To quickly find out the parameter, SD of which
would have exceeded the Priority II value (refer table 1), the parameter
wise index is scanned. ALI value is the lowest (42). It comes from
the AL1L and AL1R values (refer explanation of area marked D

50
 above). The AL1L and AL1R values are 3.6 and 3.4 mm respectively,
which are more than the priority II value of alignment on 7.2 m chord
of 2.7 mm. Checking SD values of other parameters we see that
value of AL2L and AL2R have value of 5.4 and 5.1 mm respectively
are also more than the stipulated value of 5.0 mm (priority II value on
9.6 m chord. But the values of AL1L, AL1R, AL2L and AL2R, all are
less than priority I values, so the marking in MI column would be **.
We are looking at the speed band up to 110 kmph band. SD values
of remaining other 7 parameters are less than priority II values.
These results are supposed to be used for identifying blocks with
bad track parameters so that the spot maintenance decisions can
be taken.
Area marked H to L with headings as PEAK DISTRIBUTION and
PEAK INFORMATION are for per kilometer and not for block of
200m.
4.2.6 Area marked H– This area named SHORT CHORD gives number of
locations in a kilometer (not the block of 200 m – from here onwards
we would be talking about a kilometer and not the block of 200 m)
where the defect for particular parameter exceed the category A value.
The category wise values for different parameters are given in Table
3.4 in chapter. All these are on short chords/base and that is why the
parameters are named as UN1L, UN1R, TW1 and AL1L and AL1R.
If we see parameter AL1L, there are 56 locations in this kilometer
where alignment on 3.6 m chord exceed the A category limit of 3
mm. Out of these 56 locations, at 22 locations the values have
exceeded category B limit of 5 mm. Even out of these 22 locations,
at 8 locations the values have exceeded category C limit of 7 mm.
And again out of these 8 locations, at 2 locations the values have
exceeded even the D category limit of 10 mm. Here this AL1L column
has been intentionally taken as it has exceedences in all categories.
At the bottom of all the columns is categorization like B2 for UN1L,
B2 for UN2L, and D32 for TW1 and so on is indicated. There are
three rules to do this categorization. There are two parts to it – the
letter A, B, C or D and the suffix in the form of number 2, 32 etc.
(i) If number of peaks exceed 10 (i.e. if they are 11 or above), track
is designated as next category. Say ALIR in annexure-II – Peaks
exceeding category A are more than 10, so are for category B
and the track is designated with category letter C.

51
 (ii) For the suffix we have look only at the number of peaks
exceeding B category limits. Thus if you see all the
categorization B2, B2, D32, B2, C22 and C13 for UN1L, UN1R,
TW1, G, AL1L and AL1R respectively in area marked H, all the
numbers can be seen against the row for BAND B (marked by
a rectangle in Annexure-II).
(iii) For Alignment and gauge the highest category is C and not D
(even though no. of peaks exceeding D category are also given
in Annexure-II – TRC results). That means, had the peaks
exceeding ‘C’ category for AL1L been 12 instead of 8 still the
category would have been C22 and not D22.
4.2.7 Area marked J – This area gives peak exceedences for Long Chords
and the bands are designated as W, X, Y and Z corresponding to A,
B, C and D for short chords. No categorization like W4 etc. is done.
The reason for this as that by the time long chords came into use,
the TGI based system was getting introduced and A, B, C and D
categorization had lost its significance, so no need was felt for doing
categorization on long chord.
4.2.8 Area marked K – Let us start by a slice of history to understand the
results given in this area. The maximum speed prior to 1970’s was
100 kmph on the Indian Railways. Extensive field trials and detailed
studies were done in the year 1969-70 on Delhi-Howrah Rajdhani
route to arrive at the permissible track and rolling stock tolerances,
if the trains are to run at speeds of 120 kmph. It was called poor
man’s high speed as the mandate was to find out the feasibility of
increasing the speed without incurring any substantial expenditure
on track/coaches/signaling system. The report was issued by RDSO
by the name CIVIL AND MECHANICAL ENGINEERING REPORT,
which was in 2 volumes. Volume I pertained to track and Volume II to
Mechanical department. So we talk of C & M 1 tolerances as specified
in the volume 1 of the report, whenever the speed of the section is to
be raised beyond 105kmph. A certificate to this effect is to be
furnished to PCE/CRS that the track is being maintained to C & M 1
tolerances.
These tolerances were later incorporated in the Indian Railway
Permanent Way Manual (IRPWM) through correction slip No. 45 in
paragraph 607(2) with minor modification in value of unevenness

52
 defect. The tolerances are given in table 4.2 below.
The values in column (3) in the table below are called Low Level
Band and column (4) are called as High Level Band and denoted by
abbreviations LL and HL in area marked K in the Annexure-II.
Number of locations, where the defects for respective parameters
exceed the LL and HL limits, are counted and displayed here. All
these defects are on short chord and chord for unevenness used
was 3.5m, alignment 7.5 m and twist was on the base of 3.6 m.
Table 4.2 – Peak Tolerancesas per clause607 (2) of IRPWM
(Speed Band 100-140 kmph)
Parameter Chord/ Different Bands
Base (m) Permitted Permitted at
Generally –LL Isolated Locati-
Band ons - HL Band
(1) (2) (3) (4)
Unevenness 3.5 10 mm 15 mm
Twist 3.5
a) On straight & 2 mm/m 3.5 mm/m
circular portion
b) On transition 1 mm/m 2.1 mm/m
Alignment 7.5
a) On straight 5 mm 10 mm
b) On curve +5 mm on avg. +7 mm on
versine avg. versine
c) Chord to chord 10 mm -
variation
Gauge No value specified. The maximum limit for tight
and slack gauge should as indicated in
paragraph 224 (2) (e) of IRPWM

53
 If we see the values for AL1L – at 19 locations in that kilometer the
value of alignment defect on 7.5 m chord exceeds 5 mm and out of
those 19 at three locations it exceeds 10 mm. Whenever the number
of peaks exceed permissible limits ‘#’ is marked against that
parameter, as can be seen in Annexure-II.
This particular kilometer is not being maintained to IRPWM 607(2)
tolerances as not more than 10 peaks are allowed to cross LL band
and no peak is allowed to cross HL band. These results are hardly
used (like the results of Peak Distribution – A, B, C and D
categorization) for most of the maintenance decisions. But these
are still used for taking decisions for raising the speeds of the trains
in terms of Policy Circular No. 6 for sanction of speed of nominated
trains on specific routes in terms of Railway Board’s letter No.92/
CEDO/SR/4/O/Pt .I dated 09-07-2010 and Policy Circular No. 7 for
opening of section and sanction of sectional speed Indian Railways
in terms of Railway Board’s letter No.2011/CEDO/SR/15/O dated
08-10-2012
4.2.9 Area marked L – This area named as PEAK INFORMATION gives
the isolated bad points like we note down in engine/last vehicle
inspection but more objectively and for different parameters on long
as well as short chords. The limiting value above which these defects
are listed can be set by the operator in the beginning of the TRC run.
Now if we see the 1st value against UN1R – It is -12/280, which
means that there is a defect of unevenness on right rail of 12mm on
a chord of 3.6 m at a distance of 280 m from the start of the kilometer.
–ve sign indicates that it is a dip. The worst among these are
attended by deploying gang for the same.
Vertical acceleration peaks and Lateral acceleration peaks are also
shown, denoted as VACC and LACC. The positive and negative notations
are explained in table 4.3 below.

54
 Table 4.3 – Sign convention for the Peak Information
Parameter Sign convention
Positive Negative
Unevenness Upwards Downwards
Alignment Away from the center of track Towards the center of track
Gauge Slack (Wide) Tight (Narrow)
Accelerations
(a) Vertical Upwards Downwards
(b) Lateral Towards right Towards Left

IMPORTANCE OF CHORD LENGTHS

Mid chord versines are used for measurement of unevenness
and alignment. Results are also reported on two chords as
explained above. Tracks on IR are used by slow speed stiff
suspension freight stock as also by soft suspension passenger
stock. A single chord system is not able to cover the entire
waveband of track defects that affect the dynamics of different
kinds of rolling stocks.
As broached in chapter 1, the wavelength of the defect in track
will induce vibration of a particular frequency in the rolling stock,
which will depend on the speed of the vehicle. If that frequency of
vibration matches with the natural frequency of that particular
rolling stock, the stock will exhibit wild movements (phenomenon
of resonance). Take a simplistic example of rail joint (unevenness
defect) at every 13 m in old fish plated track with a rolling stock
moving at a speed of 15.6 m/sec (56 kmph).
The relationship between frequency (f) speed (V) and wave length
(λ) is given by the equation, =V/λ . The frequency of vibration will
thus be 1.2 cycles/second (15.6/13). Natural frequency of vibration
of ICF coach is 1.2. ICF coach will thus experience wild
movements as compared to other stock at this speed of 56 Kmph.

55
Rolling Stock Natural Frequency Speed Critical
(Cycles/second) Kmph m/s Wavelength (m)
CRT Wagon 4.0 to 2.5 75 20.8 5.2 to 8.3
Box N Wagon 3.5 75 20.8 5.95
WDM2 Loco 2.4 110 30.6 12.75
ICF coach 1.2 130 36.1 30
Future Coach 1.0 160 44.4 44.4
Thus, as the speed increases and suspension becomes softer,
the chords of relevance become longer. So for these the defects
on long chord have to be attended. We should, therefore, graduate
to longer chords of 30-35 meter for Rajdhani routes.

56
 Chapter 5

OSCILLATION MONITORING SYSTEM

5.1 What is it and why we need this?

Oscillation monitoring system or OMS for short is a quick way of
judging the riding quality of the track and identifying localized track
defects. It measures the vehicle response, as excited by the track
defects, through measurement of the vertical and lateral
accelerations, as felt on the coach or locomotive/engine floor. Major
advantage of this over TRC or Oscillograph car is its portability,
simplicity of operation and also that of results. To have a quick
assessment of track in the intermediate period between TRC runs,
we need some means. OMS provides this facility. In a way we can
call it an objective engine (foot–plate) or last vehicle (brake van)
inspection. The equipment consists of portable accelerometer used
for oscillation monitoring with transducers to convert the oscillations
to electrical signals which can be recorded electronically and
processed. The box like unit has a mass “m” attached to a thin
flexible plate whose other end is firmly fixed in casing of
accelerometer. Space within casing is filled with damping fluid which
is normally silicon liquid. Flexible plate has strain gauges to sense
deflection of mass “m” which is indicator of acceleration. Schematic
diagram is shown in figure 5.1. The output can be printer based in
old machines and laptop based in new machines. Old machines
also had small memory capacity to hold data for 600-700 km of
normal quality track. It is placed on the pivot location (so as to have
same location and comparability of results) in the loco or last vehicle.

Figure 5.1 – Schematic arrangement of OMS architecture

57
5.2 Salient Features of OMS
Portable compact unit with 18 Kg weight (excluding laptop)
Battery operated: 8 hours operations with fully charged battery.
Possible to operate with 24V DC, coach battery, 110V/220V AC
supply.
Built in amplifier : No messy connections
Real Time digital output
Can be operated in tacho as well as non-tacho Mode
In tacho mode there is an option to take route data from route
tape
Can store data during run in battery backed RAM
Rugged - Does not require air conditioning
5.2.1 Latest OMS equipment are PC (Laptop) based (PC-OMS) and the
details can be seen in Version 3 of RDSO specification No.RDSO/
Res./003/2003/March 2012. These systems give a better data
storage, analysis and display features. Data can be viewed in real
time on the Laptop like TRC.
5.3 Recording of Defects
Lateral and vertical acceleration peaks exceeding the following
values are considered for track quality assessment.
Table 5.1 – Threshold values of OMS Peaks
Broad Gauge Acceleration Peaks
1. A & B routes and on routes above > 0.15 g
110 Kmph
2. Other routes with speeds up to > 0.20 g
110 Kmph
Meter Gauge
1. All routes > 0.20 g

5.4 Classification of Track Quality

5.4.1 To classify a continuous section's (PWI's jurisdiction/sub-division/

58
 division) track quality, the following criteria is to be used (average
number of peaks per km)
Table 5.2 – Classification of track quality based on Peaks/Km
Peaks per Km Very Good Good Average
1. High speed < 1.0 1-2 >2
2. Others < 1.5 1.5 – 3.0 >3

5.4.2 However, if the average number of peaks of vertical and lateral

accelerations exceeding 0.30g is more than 0.25 per km or more
than one in any particular kilometer, the track will need attention.
5.4.3 At locations where peaks of lateral and vertical accelerations exceed
0.35g, the track will have to be attended to urgently.
5.5 Frequency of Recording
OMS is to run as per the following frequency.This schedule is only a
guideline. Principal Chief Engineers may vary it, depending upon
the availability of instruments and its use. A, B and C routes are to be
compulsorily covered once a month and other routes can be covered
as per capacity and need.
Table 5.3 – Frequency of OMS recording
Broad Gauge
1. Speed above 100 Kmph Once every month
2. Others Once every two month
Meter Gauge
1. Speed above 75 Kmph Once every month
2. Others Once every two month

59
5.6 Non-recording speed for OMS
Recording by OMS done at the following speed shall be treated as
non-recorded
Table 5.4 – Non-recording speed limits for OMS
Broad Gauge
1. “A” Routes Less than 75 Kmph
2. “B” Routes Less than 65 Kmph
3. Other Routes Less than 75% of maximum
speed or 60 Kmph whichever is less
Meter Gauge
1. “Q” Routes Less than 60 Kmph
2. Other Routes Less than 55 Kmph

5.7 Sperling’s Ride Index

OMS captures measurement of lateral and vertical acceleration every
10 milliseconds (100 values per second) and uses them for
calculating Sperling Ride Index (R.I.) as explained below. All the
values of accelerations so stored in a buffer, irrespective of these
being higher or lower than the threshold value, are used for
computation of RI.

Where n = no. of completed half waves (cycles)

bi = Peak value of amplitude for the ith half-wave (cm/sec2)
fi = Frequency of the ith half wave (Hz) = 1 /(2Ti); Ti = Time of ith
half cycle
F(fi) = Correction factor for the ith half wave (Function of frequency
that takes in to account the human reaction to frequency of
acceleration)
Correction factors for various frequency values are as follows:

60
 For vertical mode For lateral mode
0 for f < 0.5 Hz 0 for f < 0.5 Hz
0.325 f for 0.5 < f < 5.4 Hz
2
0.8 f2 for 0.5 < f < 5.4 Hz
400/f2 for 5.4 < f = < 20.0 Hz 650/f2 for 5.4 < f = < 20.0 Hz
1 for > 20 Hz 1 for > 20 Hz

In Time Mode Ride Index of all blocks are calculated after completion
of a kilometer indicated by entering of Km, on the basis of elapsed
time read, by dividing it in five blocks. In Tacho Mode RI is calculated
for every block of 200m. In both cases RI of all the blocks are printed
after completion of KM. In case of Tacho mode number of blocks
can be less or more than 5, depending upon the length of the
kilometer. However, in case of time mode it will be assumed that the
km is of 1000m and the number of blocks will be 5. Value of RI
generally varies from 1 to 5. Interpretation of RI for riding is as under.
Table 5.5 – Interpretation of RI values
Ride Index (RI) Interpretation Ride Index (RI) Interpretation
1.0 Very good 3.5 Just satisfactory
1.5 Almost very good 4.0 Tolerable
2.0 Good 4.5 Not tolerable
2.5 Almost good 5.0 Dangerous in service
3.0 Satisfactory

Criteria Committee has recommended Ride Index values for different

rolling stocks in use on IR which is as under
Table 5.6 – Limiting RI values as per Criteria committee
Type of Vehicle Preferable Limiting
Diesel and electric Locomotives 3.75 4.00
Carriages 3.25 3.50
Wagons 4.25 4.50

61
5.8 Important points
In addition to track geometry, ride quality depends on
Speed of the vehicle and
Vehicle Characteristics
To eliminate above effects, successive recordings should be done
preferably using same coach in same train in same direction (up or
down) of movement. Trained operator should be used for recording.
OMS should be sent to RDSO once in eight months or whenever in
doubt for calibration check.

TYPICAL OUT PUT OF OMS

04,11,88 Date (DD/MM/YR)
12.56.16 Time (Hr/Min/Sec) 24 hr format
1251 0000 K Starting location in KM
0100 0.15 0.15 G Max. speed, V accn. Limit, L accn limit
2.50 2.50 R Max. V RI limit and L RI limit
1251 0092 K Location in KM, Distance from last KM
102 0.16 0.04 G Speed, Vert. accn., Lat. accn. in ’g’
1251 0320 K
089 0.19 0.07 G
1251 0917 K
095 0.06 0.17 G
1251 Km switch pressed
092 Average speed during last KM
01 2.93 2.33 R Block No., Vert. RI, Lat. RI
02 2.91 2.63 R
03 3.14 2.54 R
04 3.02 3.30 R
05 2.97 2.83 R

62
 Chapter 6

MAINTENANCE DECISIONS BASED ON TGI AND RELATED ISSUES

6.1 Significance of Results

Track recording car is one of the three mechanized track recording
systems, which gives us most extensive information of the track
health. The other two systems viz. Oscillation Monitoring System
(OMS) and Oscillograph Car give only the ride comfort parameters
– the lateral and vertical accelerations and the Ride Indices.
Thus these results indicate the combined effect of the track
parameters and the vehicle parameters as a function of speed.
These parameters are dependent on speed of the train during
recording. For this reason run below a speed of 75 Kmph is
considered non-recording for OMS on sections having speed of
110 kmph. Without a minimum specified speed these
parameters are meaningless as they will not give any meaningful
values – rather very good values. They do not give any information
about the parameters of track.
It is only TRC, which measures track parameters; directly as well
as indirectly, and gives various indices and defects purely for the
track. Thus for a civil engineer concerned about the track, these
are of immediate use. TRC can run even at slow speed of 20
kmph and still give meaningful results pertaining to track
parameters. Practically track parameters are independent of
speed, though in practice run at higher speed due to higher
dynamic forces is likely to cause some minor variation in the
track parameters. In fact TRC also gives the acceleration peaks
felt as well as the Ride Index values even for low recording speed
but they again are meaningless at such low speeds. But the
track parameters viz. Gauge, Alignment, Unevenness and Twist
are important for us to take the maintenance decisions.
Instructions exist for maintenance of track based on the TRC
reports. In this article we will list out and elaborate on these
instructions and also highlight certain related issues.
6.2 Maintenance decisions and TGI
In the previous chapter different results given by TRC for a typical

63
 kilometer were explained. There are three kinds of maintenance
needs for the track as far as the track geometry defects due to
general deterioration is concerned. Identifying and reporting of
these are the objective of any track monitoring system. These
are listed and explained below also discussing the issues
involved.
1) Individual isolated track defects (Worst Peak attention)
2) Bad stretches of 200 meters or more needing attention (Spot
attention)
3) Through packing (end to end) of the entire section of PWI
(Through attention)
6.2.1 Individual Isolated Point Defects(Worst Peak attention)– This
is based on the worst peak locations given in TRC results – 10
worst peaks in case of older TRC machines and peaks exceeding
preset limits in case of new TRCs. This system is satisfactory
and needs no further elaboration beyond what is already
explained in Chapter 4. The only difficulty faced is in locating
these defects in the field. At times the correlation of the kilometers/
chainages given by the TRC is difficult, also in cases where
route data is in the memory of the TRC. Due to errors in marking
the starting point and/or TRC moving a larger distance when
travelling through loop(s) at certain stations, the location given
by TRC may not be correct and in that case correlating the well-
known defects in the field can help straighten out the things.
Alternatively one can look at the TRC results and try to correlate
the known bad spots with it and try to find out the mismatch in the
length, which we may crudely call the “Zero Error”.
6.2.2 Bad stretches of 200 meters or more needing through attention
(Spot Attention)
6.2.2.1 Prior to the SD based system, only peak based system was
prevalent, primarily because the microprocessors were not there
to do multitude of calculations required for the arriving at the SD
values. Thousands of computations are required to be done per
second in the modern TRC to give the SD results, what you see
in Annexure-II. In peak based system quality of one Km of track
was worked out. It is always desirable to know the quality of even
smaller units of track to attend only those smaller stretches which

64
 need attention, rather than the entire kilometer. This became
possible with the increase in the computing power and the quality
of each block of 200m is reported in the present TRCs.
6.2.2.2 Track monitoring on standard deviation based system was
introduced on wide scale on Indian Railways as approved by
Railway Board vide letter No. 94/Tr. III/TK/23 dated 30.12.1996.
This was based on RDSO’s recommendations vide letter No.
TM/IM/TRC/Tech. (Vol. II) dated 18.9.96. The recommendation of
RDSO was based on Technical Report No.TM-11 (see box for TM
reports in Chapter 4).
6.2.2.3 Vide RDSOs recommendation (letter dated 18.9.96) paragraph
2.3, the maintenance requirements of the track were to be based
on the standard deviation values of the 4 parameters viz.
unevenness, twist, alignment and gauge in comparison to the
respective SD values specified for urgent maintenance for long
chords as well the short chords. There are 11 values of SD for
each block of 200m given in the TRC results.
(i) Alignment – 4 SD values namely AL1L, AL1R, AL2L and
AL2R – Alignment on short chord for left as well as right rail
for short chord (7.2 m – designated 1) and long chord (9.6
m – designated 2).
(ii) Unevenness – 4 SD values namely UL1L, UL1R, UL2L and
UL2R; in a similar fashion on 3.6 m and 9.6 m chords
(iii) Twist – 2 SD values on 3.6 m base and 4.8 m base TW1 and
TW2
(iv) Gauge – 1 SD value G
The monitoring by the field officials was to be done by listing out
the blocks (200 m) having higher value of SD. For this purpose
a term called “Normalized SD Value” was coined which was the
ratio (r/ru) of measured SD value (r) and the urgent maintenance
SD value (also called priority 1 value) (ru) for a particular parameter.
Thus the blocks were to be listed in the descending order of the
normalized SD value so as to prioritize the track maintenance
needs. Higher the normalized value approaching 1, nearer is the
track condition to urgent maintenance limits. Normalized value
of 1 of any parameter indicates that the SD value of that parameter
has reached urgent maintenance SD value.

65
6.2.2.4 In the letter issued by Railway Board (dated 30.12.96) for
implementation of the SD based monitoring system; the
maintenance requirement was linked with TGI value rather than
with the individual SD values or the individual indices as explained
below. These were the instructions to identify and list out the
blocks requiring spot attention.
Table 6.1 – Guidelines for Spot Attention
Band width of TGI Maintenance Requirement
80 and above No maintenance Required
50 to 80 Need based maintenance
36 to 50 Planned maintenance
Below 36 Urgent maintenance

It should be borne in mind very clearly that these instructions are

for spot attention and not for through tamping of the section of
PWI.
6.2.2.5 It happens many a times that the TGI value is more than 70 but
the index value of one of the parameters is very bad; may be
below 36 also (meaning SD value is higher than the value
stipulated for the urgent maintenance limits). This situation is
also envisaged in RDSO report No.TM-11.
Though the Railway Board’s letter mentioned that within each
TGI band individual parameters should be examined and studied,
to reach a stage of predictive maintenance, it was never put into
practice. Thus the importance of working out the maintenance
needs based on individual SD parameters was never put in
practice. People got stuck up looking at the TGI value of the blocks.
TGI value is a weighted average and gives disproportionate
weightage to the alignment parameter and vitiates our opinion if
we try to base our maintenance needs on it. It is not to say that
giving higher weightage to alignment parameter is wrong. It has
its utility when we want to know the passenger comfort levels
rather than the track maintenance needs. As a result, what is
generally perceived in the minds of the field engineers is that the
TGI value is more important than the SD values/indices of
individual parameters. This perception is required to be corrected

66
 in the minds of the field Engineers. This can be either through
education in this aspect and/or issue of revised instructions for
taking spot maintenance decisions highlighting the importance
of the SD values/indices for the individual parameter. The
parameter wise indices (which are worked out from SD values)
are also not recommended as these take into account the SD
values only on one of the 2 chords (9.6 m for unevenness, 7.2 for
alignment and 3.6 m for twist). So SD based system is desirable
for choosing the bad stretches based on higher SD values
(poorer track).
6.2.2.6 Now the question arises as to which SD value of which chord is
to be given importance – short chord or long chord? Again we
refer to TM-11, which mentions that short chords are pertinent for
freight wagons and long chords are more pertinent for coaching
stock and locos. What we should do is; depending on the
predominant traffic in our section, we should prioritize the bad
spots (200 m lengths) either on short chord or long chord. In
case the section is having balanced mix of the traffic, we should
take spots indicated by long chord as well as short chord SD
values. The offline processing software given by RDSO for each
class of TRC ((i) TRC Nos. 225 and 2500, (ii) 7965 & 7965, and
(iii) 7967 – three different classes of TRC have to use three
different software), has the facility of getting exception report of
blocks having ** or *** marking in the column MI of TRC results.
Track cell in HQ will be having the copy of offline processing
along with the raw data file of TRC results. Take it and you can
play with it. Some railway personnel have developed simple
programmes for converting the raw file into an excel sheet where
it is easier to deal, analyze and sort the data as you want.
6.2.2.7 Further, the Board’s letter dated 30.12.96 gives the same TGI
threshold values for maintenance needs irrespective of the
speeds of the section. That means there is no difference between
the TGI values for deciding maintenance needs for section having
speeds of 100 kmph and section having speed of 130kmph.
However, in case of SD values there is at least different set of SD
values for section with speeds up to 110 kmph and sections with
speeds between 110-130 kmph. So this process is more rational
on this technical ground also.

67
6.2.2.8 There are two more issues in connection with the reporting of
the km wise results of TRC.
(i) Reporting of peak excceedences and categorization of track into
A, B, C and D categories. This information is of use neither to the
higher officials nor to the field officials. It was also mentioned in
RDSO report TM-1 to delete this from the reports. It is still
continuing and it does not serve the need of either the higher
management or the lower field officials.
(ii) The tolerances as per para 607(2) are peak based tolerances
and have come from the C&M Vol. I tolerances of 1970 vintage.
These are required for certification whenever the speeds of the
section or the trains are to be raised beyond 105 Kmph. Many
CRSs have also commented that we should have SD based
criteria for the decision of increasing the speed of the section
rather than the peak based criteria {IRPWM 607(2)}. Peak based
criteria was acceptable when IR did not have the microprocessor
based TRCs. So this is one area where RDSO needs to do
some work.
6.3 Through packing of the entire section of SSE (P.Way)
Vide Railway Board’s letter dated Jan 2008 (Railway Board’s
letter No. 2007/Track III/TK/7 Pt. dated 16.01.2008), TGI based
criteria was laid down for taking decision on through packing
considering SSE(P.Way)’s (in-charge PWI) section as a unit. Prior
to these instructions the criteria for through packing was passage
of 100 GMT of traffic or 2 years, whichever was earlier on the
concrete sleeper track. The track geometry or the running quality
of the track was not a factor in this decision. In general, the
condition of an asset should always be a consideration for taking
through attention decisions. More so in case of the track, the
dynamic behavior of which is the condition of the asset and it
should be an important criteria to decide the requirement of major
maintenance inputs like through packing/tamping. This was
attempted to be corrected by these instructions of 2008.
One thing which we should be aware of is that machine packing,
apart from consuming direct resources consumes very important
and critical resource – the block. And over tamping is detrimental
to track in more ways rather than being beneficial except for

68
 temporary improvement in running. The temporariness as an
objective generally leads to inefficient use of the direct resources.
The purpose of issue of the present instructions in 2008 was to
• Discourage the unnecessary tamping of the track
• Have some objective criteria to monitor the quality of tamping
work done
These instructions were issued as a result of study done by
RDSO and published in Technical Report No.TM-115. TGI based
stipulation for tamping is reproduced below.
Table 6.2 – Guidelines for Through Attention –
Speed > 110 & < 130 kmph
S. No. INDICES TGI value
1. 60% CFD before tamping (to decide < 90
to give machine for through tamping)
2. 60% CFD after tamping (to ensure quality of > 115
through tamping) (i.e. minimum 40% of
track length after tamping should have TGI)
3. 10% CFD after tamping (to ensure > 75
quality of through tamping) (i.e. minimum
90% of track after tamping should have TGI)
Table 6.3 – Guidelines for Through Attention –
Speeds up to and less than 110 kmph
S. No. INDICES TGI value
1. 60% CFD before tamping (to decide to < 80
give machine for through tamping)
2. 60% CFD after tamping (to ensure quality of > 115
through tamping) (i.e. minimum 40% of
track length after tamping should have TGI)
3. 10% CFD after tamping (to ensure quality > 75
of through tamping) (i.e. minimum 90% of
track after tamping should have TGI)
*PWI/In-charge unit is the basic unit for the above criteria

69
 These stipulations were based on actual data of TGI values
collected from various zonal railways; pre-tamping as well as
post-tamping. The table showing the data as below has been
taken from RDSO report No.TM-115.
Table 6.4 – Pre-tamping and post-tamping TGIs –from TM-115
S. Railway Route Maximum TGI values
No. 60% CFD 60% CFD 10% CFD
before after after
tamping tamping tamping
1. SCR A 110,120 85 95 70
2. SCR B 100 90 90 65
3. WR A 120,130 75 100 75
4. WR A&B 100 82 113 80
5. NER D 110 77 90 50
6. NER D 100 77 90 50
7. SER A 105 84 102 76
8. SER B 105 84 90 65
9. SEC A 110 84 90 62
10. SEC A 105 78 93 70
11. ECR B 105 102 117 76
12. SR B 110 106 107 72
13. NCR A 110 103 111 70
14. NR B&D 50,110 81 88 54
Average Speeds>=110 kmph 91 101 70
Speeds<110 kmph 85 98 67

On perusal of this table it is seen that only one Railway i.e. ECR
Railway (S. No.11) has achieved the stipulated track area post
tamping and that too because machine tamping was done even
though it was not required to be done as per criteria. Apart from
this issue, there are certain other issues which are coming up in
implementation of the above instructions –

70
(i) One issue worth pondering is that the TGI values do not
directly indicate the maintenance needs as it is correlated
with the ride index (reflecting riding comfort) and therefore
reflect the overall condition of track from comfort point of
view. And also that tamping is not the only way to improve
the TGI value. So instead of TGI value which is weighted
average of parameters for 4 indices, value based on 4
parameters directly or having different weightages which
would indicate the maintenance needs rather than the riding
comfort should be the basis for the working out of the
maintenance needs.
(ii) Presently, the reward system is based on the TGI value so
it is natural for the field people to work for the improvement
of TGI and not on optimizing the use of the tamping machine
and use it only when necessary. The reward system will
have to be shifted to encourage the reduction of wasteful
usage of tamping machine also. The conflicting demands
between the annual targets output of tamping machines
fixed by Railway Administration and the idea of reducing
unnecessary deployment of machine have also to be
resolved.
(iii) Though the post-tamping TGI values have been stipulated,
a mechanism has to be put in place to monitor and reward
it. RDSO had recommended it in TM-115 but it was finally
not included in the letter issued to the Railways for
implementation of CFD based criteria.
(iv) There would be many PWI sections on main line routes,
where the pre-tamping TGI values would always be above
the values stipulated for through packing. That PWI will not
get the machine till the TGI value of 60% section drops to
below 90. The purpose of these instructions is not to bring
about socialism based on TGI values but to retain the good
sections as good and bring about substantial improvement
in the bad sections.
(v) The stipulations are being considered unachievable by the
railways and it is well known that only realistic achievable
targets excite people. So the post tamping TGI values to be
achieved needs revision as discussed below in point no.

71
 (vi). Even a 10 points improvement in the TGI value is quite
substantial particularly more so in a good track; so expecting
TGI improvement with one round of tamping by 25 TGI points
(35 for the sections with speeds up to 110 kmph) is
unrealistic.
(vi) The TGI which can be achieved on any section depends on
many factors like
• Maximum permissible speed of the section
• Condition of the formation
• Age and condition of track fittings including rails and sleepers
• Condition of ballast cushion, deep screening and its
resilience etc.
• Track geometry – like number of curves (in Ghat sections
etc.)
• Axle loads
• Extent of traffic mix
• Memory of the track
• Rainfall and climatic conditions
• Any other factors which may affect retention of the packing
Thus, all sections cannot be put at par in view of such large
number of factors which have a substantial impact on the TGI
value.
Relevant excerpt from RDSO report is quoted “It should be noted
that track maintenance requirements cannot be assessed
merely on the basis of track geometry values. The condition
of the track components and fitting, the ballast bed and other
local factors should be taken into account while assigning
priority for maintenance. For planned maintenance, the rate
of deterioration is an important factor. The rate of deterioration
can be assessed by building up historical record of successive
recordings.”
There is no doubt that some objective criteria have to be there
for taking up the through packing as well as to monitor TGI

72
 value improvements after the tamping.
It is therefore desirable to compare the TGI values with the historic
TGI values of the same section, instead of specifying the absolute
values; which would automatically take into account the above
factors. Thus, it is recommended that zonal railways should be
given a matrix and broad guidelines for fixing threshold TGI values
for through packing as well as TGI values to be achieved after
through packing for different sections. And the post-tamping TGI
values can be in terms of percentage of the pre-tamping values
rather than the absolute values. Not all the factors as mentioned
above can be quantified but at least they can form the basis for
working out the acceptable values.
This chapter has discussed these issues with the objective to
initiate the thought process. Some of the issues are under
consideration of RDSO and Railway Board after they were
discussed in CTE, CE(TM) and PCE’s seminar at IRICEN and
also in other forums outside IRICEN. It will make understanding
the issues involved better apart from enlightening the readers
about the subject matter. This will help in getting the diverse and
rich opinions of the experienced field engineers, who normally
are not able to contribute their thoughts, due to lack of
understanding.
6.4 Running of TRC
6.4.1 Arrangements for running track recording car
On receipt of track recording car programme from the R.D.S.O.,
the Zonal Railways should arrange for suitable power and path
for the special along with telecommunication arrangement
between the track recording car and the locomotive. The
Headquarters should advise the Divisions concerned for making
necessary arrangements to ensure that the Track Recording
Car has an uninterrupted run to reduce the non-recorded
kilometers.
6.4.2 Actual running of Track Recording Car – The Divisional Engineer,
Assistant Engineer and Permanent Way Inspector of the section
and nominated officer/staff of Headquarters office should
accompany the Track Recording run. The Track Recording Car
results vary to some extent with speed. The Divisional/

73
 Headquarters Officers accompanying the special should
therefore ensure that the Track Recording Cars are run at the
specified recording speed range of the particular car. The
desirable recording speed of a Broad Gauge car is 75 to 105
Kmph.
The recording done below the speed of 20 Kmph in the current
cars (No.7965, 7966 & 7967) are taken as “Non-recorded”. For
obtaining comparable results between successive recordings,
it is necessary to run the special at uniform speed. The track
recording car specials must run on through lines of all stations.
Recording should preferably be done during day light hours.
Before start of any run it should be ensured that the calibration of
the car has been done properly by RDSO as explained in
Chapter 9 “How TRC measures track parameters”. The charts
should be taken out and sent to Headquarters Office for obtaining
copies after the run - samples of charts obtained.
6.5 Frequency of Track Recording
The Broad Gauge routes should be monitored by TRC as per the
following frequencies (except for the routes where track recording
is to be dispensed with):-
Table 6.5 – Frequency of Track Recording Car Run
S. No. Route Frequency
(i) Routes with existing speeds above Once in 2 months
130 kmph
(ii) Routes with existing speeds above Once in 3 months
110 kmph & up to 130 kmph
(iii) Other Group ‘A’ and ‘B’ routes Once in 4 months
(iv) Group ‘C’, ‘D’ and ‘D Special’ routes Once in 6 months
(v) Group ‘E’ and ‘E Special’ routes Once in 12 months

6.6 Oscillograph Car

6.6.1 Brief description of the car: – The main equipment in this car is
an accelerometer, which consists essentially of a mass attached
to one end of thin flexible plate, the other end of the plate being
fixed firmly to the casing of the accelerometer. The space within

74
 the accelerometer casing is filled with a damping fluid (usually
silicon liquid). The cantilever plate is strain gauged so that the
deflection of the mass ‘m’ is sensed. This is calibrated to indicate
the acceleration to which the mass is subjected. This
acceleration is recorded in the form of an accelerogram. Thus
the vertical and lateral acceleration on any part of the vehicle
where the accelerometer is installed can be recorded. In track
monitoring runs the accelerations at the loco cab floor are
recorded by keeping the accelerometer as close to bogie pivot
as possible.
6.6.2 Details of recording: – the following parameters are recorded in
the Oscillograph car runs: – (a) Vertical acceleration of loco cab.
(b) Lateral acceleration of loco cab
In addition to the above, timing marks-one for every second-
telegraph and km. marks and even marks such as station
buildings, bridges etc., are also recorded on the chart. The primary
spring deflections left and right are also recorded.
6.6.3 Frequency of recording – These are used to monitor only A routes.
These cars are run once in six months to assess the riding
quality of the track as distinct from actual track geometry recorded
by the Track Recording Cars which have different frequencies of
runs. Recording is done at the maximum sanctioned speed of
the section.
6.6.4 Analysis of Oscillograph results and its interpretation
The results obtained from the Oscillograph car are analyzed for
the vertical and lateral acceleration.
The analysis of records is done on the following lines:–
(a) All peaks are to be measured from the base or zero line i.e.
no allowance is to be given when the trace shifts because
of cant deficiency on curves. However, for the purpose of
“ride index” calculations, lateral acceleration peaks are to
be measured from the shifted base line on curves.
(b) The vertical and lateral accelerations above the threshold
values are separately counted: Threshold value of
acceleration may be taken as follows:–(i) In Loco/
Engine Cab floor – The threshold value of acceleration in

75
 vertical mode is taken as 0.20 g. for all locos (Diesel and
Electric).The threshold value of acceleration in lateral mode
is taken as 0.20g for diesel and electric locos with double
stage suspension (i.e., for WDM-3, WDM-4, WDM-1, WDM-
2/3 and WAG-2) with the exception of WDM-2 modified with
single stage suspension. In case of other locos, with single
stage suspension (i.e.,WDM-2 std., WAM-1, WAM-4, WAG-
1,WAG-3,WAG-4,WCG-2, WCAM, WCM, WCM-2, WCM-4 and
WCM-5) the threshold value may be taken as 0.30g.
(ii) On Passenger Coach Floor: – The threshold value of
acceleration for both vertical and lateral modes shall be
taken as 0.15g.
(iii) The analysis is done kilometer-wise and after counting the
peaks above threshold value for the particular locomotive
results are given under the following heads – Station Yards,
Other than Station Yards (Isolated locations), Active
continuous stretches, Speed grouping table is also
prepared.
6.7 Procedure in the event of delay TRC/Oscillograph car runs
TRC and Oscillograph runs at the frequency as above. There are
occasions when the duration between the two runs exceeds the
stipulated duration. There are instructions contained in the letter
from CRSs as listed below about the action to be taken to certify
the track in such an eventuality for routes on which Rajdhani
Expresses run. Similarly the period of 3 months mentioned in
the instructions would be replaced by 2 months in sections having
speeds>130 kmph.
• CRS/Eastern Circle’s letter No. 549 dated 08/07/1980
• CRS/Western Circle’s letter No. 30-5/20/886 dated 05/08/1980
• CRS/Central Circle’s letter No. 374 dated 20/05/1988
The procedure to be followed in the event of TRC car delayed
beyond 3 months and Oscillograph car delayed beyond 6 months
is given below.
i) TRC/Oscillograph fails to run within 3rd/6th months of the date
of the last run – Locomotive inspection is to be done on
Rajdhani Express by J.A. Grade officer nominated by CE/CTE.

76
 A monitoring report and a safety certificate are to be issued.
The CE/CTE to issue an order permitting the continuation of
speed of Rajdhani Express for the 4th month. A copy of the
certificate to be sent to CRS for records.
ii) TRC/Oscillograph fails to run in the 4th/7th month – The above
procedure will have to be followed for the normal running of
trains in the 5th month except that the monitoring on the foot-
plate of Rajdhani Express should be done by Additional CE
instead of J.A. Grade Officer. In the present scenario it would
mean a SAG officer nominated by PCE/CTE. Issue of certificate
and sending the copy to CRS would have to be done.
iii) TRC/Oscillograph fails to run in the 5th/8th month – The above
procedure will have to be followed except that track monitoring
on the locomotive, of the Rajdhani Express should be done
by the CTE. Issue of certificate and sending the copy to CRS
would be done.
iv) TRC/Oscillograph fails to run in the 6th/9th month – In the
event of the TRC/Oscillograph not being run within 6/9months
respectively, of the date of the last run, the speed on the
Rajdhani Route on expiry of 6/9 months, should be brought
down to 110 Kmph but the Rajdhani Express itself should be
booked to run at a speed of only 100 Kmph.
Logically the Rajdhani Express or for that matter any train having
better coaches can run at a speed higher than other trains but
severe restriction has been imposed on Rajdhani Express train
to put pressure on the system for arranging the run of TRC/
Oscillograph car early.

77
 Chapter 7

TRACK TOLERANCES

7.1 Why Tolerances?

Track is a vital and costly asset for any rail transport network.
Track has to be laid to desirable standards and then maintained
not only for good riding comfort and safety but also for longer life
of the assets. Quality control of track geometry is essential while
laying new track and subsequently to ensure effectiveness of
maintenance operations. In a good system, the maintenance
operations are need based rather than periodic. For this purpose
it is necessary to lay down the limits of acceptable track
parameters so that whenever these parameters are likely to be
reached, maintenance input can be planned. Different types of
track tolerances are required to be defined, each for a particular
purpose. With the introduction of TMS systems, the role of
tolerances will also be seen in predicting the deterioration of
track, based on the past trend (of 3 or more runs of TRC) and
planning of deployment of track machines.
Maintenance practices take time to grow in the organization, but
once they take roots like trees it is not easy to change (uproot)
them. So, one has to be thoughtful in deciding them as they have
far reaching consequences. There are many tolerances prevalent
in different Railway systems in the world. Effort has been made
to have a look at those and see what is prescribed on Indian
Railways.
7.2 Track Maintenance practices on Indian Railways
Indian Railways Network now comprises mostly of Broad Gauge
(BG) Track and whatever Meter Gauge and Narrow Gauge lines
are remaining are unimportant lines. So the new tolerances
specified are for Broad Gauge only. The discussion in this chapter
would be restricted to BG only. Two more reasons due to which
laying of track tolerances for new track is important are -
• Laying of tracks by Contractual agencies
• Deploying of track machines after track is brought to certain
level of quality by manual methods

78
7.3 Type of Tolerances
Right from the time new track is laid till it remains in service; it
lives through numerous cycles of deterioration and maintenance
inputs. The deterioration of the track geometry takes place due to
impact of the wheels of rolling stock on the track under repeated
loading. The maintenance inputs can be in the form of attention
by track machines and also by renewal of components. The
condition of the track deterioration with respect to time is depicted
in the graph form below, which also shows time on X-axis, the
quality of track on Y-axis and various types of track tolerances.

Figure 7.1 – Track quality v/s time showing different kinds of

tolerances
This graph shows the deterioration of track with time as shown
in downward pointing sloping red curved lines. Periodic
maintenance inputs are given as shown as upward pointing
vertical green lines in the above figure.
7.3.1 New Track Tolerances
These are the track tolerances that are to be attained in new
track laying. These are for new lines, gauge conversions as well
as for Complete Track Renewal works in open line. These are
the best track standards in terms of track quality as also seen in
figure 7.1. Track relaying, whether done manually or using
machines, are increasingly being done by the contractual
agencies. The quality of the work has to be compared to the
stipulated benchmarks before final payments are released.
It is understandable that a nicely laid track results in lesser

79
 maintenance inputs and longer life of track assets besides
resulting in savings in fuel consumption. So the tolerances are
fixed on these considerations. The second consideration is that
the laid down tolerances should be practically achievable.
New track tolerances had been deliberated in detail in the 59th
TSC meeting held in 1984. These have been fixed in peak terms
and incorporated in Indian Railway Permanent Way Manual
(IRPWM) in clause 316 and is reproduced below.
Table 7.1 – New Track tolerances as per clause 316 of IRPWM
Gauge Sleeper to sleeper variation + 2mm
Expansion Gap Average gap over 20 readings + 2mm
Joints Low not permitted, High joint < 2mm
Squareness of joints in straight Upto + 10mm
Sleeper spacing With respect to the oretical spacing + 20mm
Cross level measured on every 4 sleeper
th
+ 3mm
Alignment On 10m Chord on straight + 2mm
Versines on curves From theoretical versine on + 5 mm
on 20m chord Curves > 600 m radius
From theoretical versine on + 10 mm
Curves < 600m radius
Longitudinal Level Variation over approved L-level up to + 50mm
The tolerances on new track on curves are also given for BG in
clause 403 of IRPWM.
Table 7.2 – Gauge values for New Track on curves
S. No. Radius in Meters Gauge
1. Straight including curves with radius -5mm to +3mm.
up to 350 M.
2. For curves of radius less than 350 M Up to +10mm.

Slack gauge is provided using slack gauge sleepers

manufactured for curves specifically up to 10 mm slack gauge in
steps of 2 mm to cater for requirements of transition and curves
of different radii.

80
 SD based tolerance for new track were deliberated as item No.786
of 59th TSC and are given in table 3.6. If the track is maintained to
new track standards in terms of SD based tolerances, the value
of the indices for four parameters will come to 100 as we have
seen earlier. These are reproduced below for review and save
the reader from turning back the pages of the book.
Table 7.3 – SD based tolerances for New Track
Parameter Chord/ Base (M) SD value for new track (mm)
Unevenness 3.6 1.20
9.6 2.50*
Twist 3.6 1.75*
4.8 2.50
Alignment 7.2 1.5*
9.6 2.50
Gauge - 1.0*
*Values in bold and marked thus used for calculation of parameter
wise indices UNI, GI, TWI and ALI
7.3.2 Maintenance Tolerances
These are next best tolerances after new track tolerances. These
are the standards of track which are to be achieved immediately
after maintenance cycle. In the time v/s quality graph, the actual
level of track quality attainable after each maintenance cycle goes
on reducing as characterized by the downward sloping yellow
straight line. This is so as the track components wear out with
time.
The basic objective of the maintenance tolerances is to ensure
that the maintenance operation has served its intended purpose.
These standards would in reality depend on the age of the track
structure, cumulative GMT carried, method of maintenance etc.;
but presently the tolerances fixed do not take these factors into
account.

81
 Table 3.1 of Indian Railway Track Machine Manual (IRTMM) lays
down single set of values irrespective of the factors mentioned
in preceding paragraph.
Table 7.4 – Maintenance tolerances as per IRTMM
Parameter Not more than 10 peaks Any peak excee-
per km to exceed this value ding this value
Alignment ±4mm ±6mm
Cross Level ±6mm –
Unevenness 6mm 10mm
Note- (i) At least 4 stretches of 25 sleepers per km of tamped
track should be recorded
(ii) The versine and SE of a curve for at least 10 stations should
be recorded

In a widely varying track structure, maintenance practice and traffic

density conditions as on Indian Railways, it is not advisable to
lay a single set of maintenance tolerances. It is therefore; felt that
on IR, laying of track tolerances is better left to maintenance
engineer’s discretion (read Zonal Railways). However, RDSO
can issue broad guidelines, depending on the various factors.
Since maintenance on most of the routes on IR are by track
machines, the tolerances so fixed will help in checking the efficacy
of maintenance operations; and if these are not achieved, then
to find out the reasons for the same and take decisions for
renewal of track components.
These have also been stipulated for section of SSE as a whole
in 2008, as explained in Chapter 6 – table 6.3, portion of which is
reproduced below.
Table 7.5 – Portion of table 6.2 & 6.3 – Maintenance tolerances
1. 60% CFD after tamping (to ensure quality of > 115
through tamping) (i.e. minimum 40% of track length
after tamping should have TGI)
2. 10% CFD after tamping (to ensure quality of > 75
through tamping) (i.e. minimum 90% of track after
tamping should have TGI)

82
 These, though, are not being implemented as discussed in that
Chapter, and are under review.
7.3.3 Index Tolerances
The basic objective of fixing this tolerance is to take care of the
time that may elapse between planning of track maintenance
and actual maintenance operation materializing as per the
existing systems and practices of deployment of machines. This
will ensure that the service tolerances (discussed in 7.3.4) are
not exceeded. These tolerances are fixed based on the rate of
track deterioration and the delay that can occur in sending a
machine for maintenance. It thus goes without saying that these
tolerances are essential on all machine maintained sections.
SD based index tolerances were fixed as per item 786 of 59th
TSC as reproduced below. These were earlier called SD values
for planned maintenance and since 2007; these are known as
Priority-II SD values. In Km-wise TRC results, blocks where values
are exceeded these are marked by ** under the column heading
MI (Maintenance Instructions). These are different for two speed
band - up to 110 kmph and > 110 up to 130 Kmph.
Table 7.6 – SD based Index Tolerances (Priority-II)
Parameter Chord/ Up to 110 >110<=130
Base (M) KMPH KMPH
(<105)* mm (>105)* mm
Unevenness 3.6 2.5 2.3
9.6 6.5 5.1
Twist 3.6 3.8 3.4
4.8 4.2 3.8
Alignment 7.2 2.7 2.3
9.6 5.0 3.3
Gauge - 2.5 1.8
* Threshold limit was 105 kmph earlier

Index tolerances are very much in use in those countries which

have centralized maintenance system and use a central

83
 computer to deploy track machines. This central planning system
directs the machine to various spots keeping in view the urgency
of maintenance, taking into account the delay that can be accepted
while also ensuring that the idling (non-working) time of the
machines is least. Indian Railways is not having such system,
though the first steps have been taken in the form of track
degradation model in TMS module.
In manually maintained track, since there is not much delay in
deployment of gangs, index tolerances can be same as the
service tolerances.
7.3.4 Service Tolerances
These are the limiting value of the track parameters, up to which
the track geometry can be allowed to deteriorate without the ride
quality falling below the stipulated ride index values. In other
words these are the limits up to which the minimum acceptable
level of riding comfort is ensured.
Frequent track maintenance is not desirable for many reasons
like being costly and consuming precious resources of traffic
block and track machines, causing damage to the consolidated
ballast bed etc. Thus lesser the maintenance, better it is. So,
these tolerances are used to identify the track needing attention.
These are, therefore, the most important tolerances for the
maintenance Engineer. These would also depend on the speed
at which the trains are operating in that section. These have
been laid down on Indian Railways in terms of peak value of
defects (for speed > 100 Kmph) and also in terms of SD values
of track parameters, for two speed bands – up to 110 kmph and
> 110 up to 130 Kmph.
These peak based tolerances given in Para 607(2) of IRPWM
can be seen in chapter 4 table 4.2 and are for the speeds greater
than 100 Kmph and up to 140 Kmph. No peak based tolerances
have been laid for speeds less than and equal to 100 Kmph. In
chapter 2, A, B, C & D categorization of track was explained;
these are not service or index tolerances and they are just some
limits specified for classification of track on quality considerations.
If we compare the tolerances given in 607(2) with A, B, C & D
limits, no close relationship is observed as seen from the
comparison in table 7.7 below.

84
 Table 7.7 – Comparison in the values in 607(2) and that of
A, B, C & D limits
Parameter Para 607(2) values of A, B, C & D limits
service tolerance (Table 3.4)
(Table 4.2)
Unevenness 10 mm (3.5 m Chord) ‘B’ 6 to 10 mm (3.6 m chord)
Alignment 5 mm (7.5 m Chord) ‘C’ 5 mm and above (7.2 m Chord)
Twist 2 mm/m (7.0 mm) (3.5 m base) ‘B’ 5.0 to 7.5 mm (3.6 m base)

Individual bad spots are thus required to be attended if the above

peak values are exceeding.
With the machine maintained sections, it is practically impossible
to attend individual bad spots. So the TRC gives bad stretches of
minimum block (200 m) length. SD based service tolerances were
fixed as per item 786 of 59th TSC as reproduced below. These were
earlier called SD values for urgent maintenance and since 2007;
these are known as Priority-I SD values. These are identified by
comparing the SD values of the individual parameters with the
respective SD based service tolerances for respective chords, and
marking them as *** in the TRC result under the heading MI in the
kilometer wise chart. These are different for two speed band - up to
110 kmph and > 110 up to 130 Kmph.
Table 7.8 – SD based Service Tolerances (Priority-I)
Parameter Chord/ Up to 110 >110<=130
Base (M) KMPH KMPH
(<105)* mm (>105)* mm
Unevenness 3.6 3.3 3.0
9.6 7.4 6.2
Twist 3.6 4.2 3.8
4.8 5.0 4.5
Alignment 7.2 3.8 3.0
9.6 6.3 4.0
Gauge - 3.6 2.6
* Threshold limit was 105 kmph earlier

85
 Then there is value of service tolerance for gauge on curves as
stipulated in clause 224(e) of IRPWM as below.
Table 7.9 – Service tolerances for gauge – 224 (e) of IRPWM
Parameter Value (mm)
On straight -6mm to +6mm
On curve with
Radius 350 m or more -6mm to +15mm
Radius less than 350 m Up to +20 mm
Note- Uniform gauge over a long stretch is desirable

Lots of problems were being faced by the field engineers after

introduction of service tolerances as per clause 607(2) in IRPWM
through correction slip No.45. In many derailment cases,
responsibility of derailment was fixed on engineering account
on account of track parameters exceeding those limits. Though
values of track parameters in clause 607(2) were laid down
based on comfort criteria, these were being treated as safety
tolerances by other departments. To clarify the issue, correction
slip No.96 to this effect was issued forming clause 607(3) of
IRPWM. This is reproduced below
607(3) of IRPWM – through correction slip No.96
“The stability of trains against derailment depends upon several
factors such as track geometry, vehicle characteristics & state of
their maintenance and speed of the particular vehicle at relevant
point of time etc. Rail wheel interaction is thus, a complex
phenomenon and therefore, safety tolerances for track alone
cannot be prescribed in isolation. With this in view, safety
tolerances for maintenance of track have not been prescribed
on Indian Railways. Each derailment case, therefore, needs
careful examination of all available evidence, in respect of Track,
Rolling Stock, speed and other factors considered relevant, to
arrive at the cause.
The provision and tolerances mentioned in Para 607(2) and
elsewhere in this Manual are with a view to maintain track
geometry for good riding comfort.”

86
 Then on macro basis service tolerance to take decision of doing
the through packing/tamping of the section is also laid down for
section of SSE as a whole in 2008, as explained in Chapter 6 –
table 6.3, portion of which is reproduced below.
Table 7.10 – Portion of table 6.2 & 6.3 – Service tolerance
for through packing decisions
S. No. INDICES Speed TGI value
1. 60% CFD before tamping > 110 Kmph < 90
(to decide to give machine for & < 130 Kmph
through tamping)
2. 60% CFD before tamping <110 Kmph < 80
(to decide to give machine for
through tamping)

Then service tolerances in broad terms have also been specified

for taking decision on realignment of curves. These are as per
CS No 90 included as Clause 421 of IRPWM. The criteria are
reproduced below and curve is taken for realignment if the values
are exceeded by more than 20% of the readings (stations).
Table 7.11 – Clause 421 of IRPWM –
Provision for realignment of curves
Speed Range Limits of station to station versine (mm)
120 Kmph and above 10 mm or 25% of the average versine
on circular curve whichever is more
Below 120 Kmph and up 15 mm or 25% of the average versine
to 80 Kmph on circular curve whichever is more
Below 80 Kmph and up 40 mm or 25% of the average versine
to 50 Kmph on circular curve whichever is more

7.3.5 Slow down Tolerances

We have seen that service tolerances are the limiting values
from the ride considerations and at no time should the track
quality fall below these values. But if the track geometry
deteriorates below service tolerance limits, to ensure the

87
 minimum level of passenger comfort, speeds will have to be
reduced. In this discussion, the reference is to passenger
services. Specifying these tolerances on high speed routes is
essential. Suppose we start operating trains at speeds of 160
Kmph, then slow down tolerances will have to be specified to
reduce the speeds to 130 Kmph. One such study was
commissioned by RDSO and a consultancy was given to M/s
Transmark Corporation of UK. They did studies and came out
with maximum permissible values of SD for respective
parameters (higher than priority or urgent maintenance limits)
for different speeds for speed bands of 10 Kmph intervals. Table
from that study is reproduced below to clearly give an idea to the
readers as to how it was proposed to be done. Since this report
did not get the approval, it was never implemented. Nonetheless
it was a logical step in right direction.
Table 7.12 – Proposed slowdown tolerances for Passenger Services
– Transmark of UK
Speed Unevenness Alignment Twist
km/h 9.6 m Chord 7.2 m Chord 3.6 m base
From To Plan Urgent Max Plan Urgent Max Plan Urgent Max
ned ned ned
0 60 5.3 7.2 9.7 3.3 4.2 5.5 3.0 3.7 4.5
60 70 5.2 7 9.4 3.2 4.2 5.4 3.0 3.6 4.4
70 80 5 6.8 9.1 3.1 4.0 5.2 2.9 3.5 4.3
80 90 4.9 6.6 8.8 3.0 3.9 5.1 2.9 3.4 4.2
90 100 4.7 6.4 8.6 2.9 3.8 4.9 2.8 3.4 4.1
100 110 4.6 6.2 8.3 2.9 3.7 4.8 2.8 3.3 4.0
110 120 4.5 6.1 8.1 2.8 3.7 4.7 2.7 3.3 3.9
120 130 4.4 5.9 7.9 2.7 3.6 4.6 2.7 3.2 3.9
130 140 3.8 5.1 6.8 2.5 3.2 4.0 2.5 2.9 3.5
140 150 3.3 4.4 5.9 2.3 2.8 3.6 2.3 2.7 3.2
150 160 2.9 3.0 5.2 2.0 2.6 3.3 2.2 2.5 3.0

88
 The concept of slowdown tolerances originated from advanced
world railways, which operate at high speeds ranging from 200
– 250 Kmph.
On Indian Railway there was one rolling stock more prone to
derailment – CRT wagons – 4 wheeler wagon. Looking at the
derailments, and number of incidences in which responsibility
was fixed on the field engineering officials, the matter was
investigated by RDSO. After the study RDSO specified slowdown
tolerances for this freight wagon. Since these could not be made
a part of IRPWM, cognizance is normally not given by other
departments in derailment cases. But these are scientifically
fixed values given in RDSO Letter No. CRA 501 dated 29-4-1983
and is reproduced below.
Table 7.13 – Parameters for slow down tolerance for CRT wagons
Speed (Kmph) Peak Value of Peak value of
unevenness on twist on 3.6
3.6 m chord (mm) m base (mm)
75 14 13
60 16 15
45 22 22
30 24 25
15 33 30

7.3.6 Safety Tolerances

Track geometry if maintained within tolerances for ride comfort,
will automatically ensure safety of vehicle as safety tolerances
will be very slack as compared to service tolerances for minimum
level of ride comfort. And as very lucidly stated in clause 607(3) of
IRPWM, and reproduced in 7.3.4 above, safety tolerances have
not been specified even on world railways. The main reason is
that rail wheel interaction being a complex phenomenon; it is not
possible to fix safety tolerance of track in isolation. Even when
new rolling stocks are introduced on IR, many speed trials are
carried out and elaborate information of behaviour is gathered.
After that service tolerances are fixed and are considered
sufficient.

89
7.4 Philosophy in fixing of Track Tolerances
All the tolerances have to be laid down keeping in view the track-
vehicle interaction phenomenon, so that high degree of correlation
between track irregularity and vehicle ride is obtained. The
parameters for unevenness and alignment are to be arrived
based on the ride index and peak accelerations in the vertical
and lateral directions respectively. Twist parameters are based
on the off-loading of the wheels. This is how it was done when
C&M-1 report was prepared and peak based tolerances were
fixed.
For the SD based system, we need the new track tolerances and
service tolerances. If we look at the actual values of SDs on
selected routes and separate out the best 5% and remaining
95% values, we will be able to fix the new track standards as the
average values of 5% best track. That is how RDSO has done it.
Index tolerances and service tolerances were fixed for the
following values of Ride Index for different kinds of rolling stock.
Other tolerances are fixed in a similar manner.
Table7.14 – Limiting RI values
Rolling Stock RI values RI values
Loco 4.00 4.25
Wagon 4.50 4.75
Coach 3.50 3.75

Fixing of right tolerances of different kinds will not only help the
field engineers to monitor the track geometry properly but also
economize the maintenance costs and increase the life of the
track.

90
 Chapter 8

OFFLINE & PRINTING SOFTWARE AND TMS FOR

REPORT GENERATION

8.1 Need for Offline printing software

Kilometer wise results of TRC run are printed out during the run. If
printouts of the results are required after the run, then software is
required to open the run data files and give output in proper form.
This software is called offline printing software. It can also print
exception reports. The offline printing software is different for different
vintage of TRC. Presently only 3 TRC’s are under operation – 7965,
7966 and 7967. Offline printing software is same for first two i.e.
7965 and 7966 and is different for 7967. Offline printing software is
available with RDSO and the Track cell at the Zonal HQ dealing with
the subject of TRC.
The reports available in offline printing software can be seen in
edited exhaustive screenshot marked as fig. 8.1 below. Since this is
old software, the drive where the run files will be kept has to be
selected. The run files necessarily have to be saved in the folder
named “processdata” in the root of that directory, which may be
external like pen drive. Run no. has to be remembered from the file
name.

91
 Figure 8.1 Edited full menu screen shot of Offline and Printing
software – 7965 & 7966
8.2 Track Management System (TRC & OMS Module)
Track Management System (TMS) is being implemented on Indian
Railways and TRC reports and analysis is one of the modules of
this. All the reports which are available in the offline printing software
and many more can be generated in TMS.

92
 Figure 8.2 Screen shot of TRC reports in TMS
There are many reports which can be printed using the above
software. It is considered prudent to mention about few of the
important reports which can be obtained by the above two softwares.
Screenshot of reports in TMS clearly demonstrate the reports.

93
S. Offline and Printing Software TMS – Reports – TRC Reports
No.
1. Kilometer wise chart which is Kilometer wise Chart
also given during the TRC run
2. Exception Report for planned Pictorial Block wise SD
and urgent limits values and color coded report
for attention to track
3. Comparing Runs CFD comparisons for up to 6
TRC runs
4. Summary report Indices Exception Report for blocks
5. Analysis of IRPWM Analysis w.r.t. para 607(2) of
IRPWM
6. Parameter wise comparison
of two runs – Pictorial in
graph with CFD % on x-axis
and SD value of parameter
like UN1L, AL1R etc.

For TMS to give reports, the raw data has to be fed into the system.
For this a separate admin login by the name “trcadmin<division
code>” has been created. The purpose is to extract the TRC raw
data file from the CD to TMS server. And since the raw data file is for
very long section, the file is broken into various sections (by the
menu command called bifurcation). This makes the file ready for
extracting various reports as shown in the screen shot at figure 8.2.
This work has to be done at divisional level.
It is desirable that readers explore the different kinds of reports
available in TMS apart from the ones depicted below.
With better quality and easy to generate reports from TMS, the offline
& Printing software would become obsolete.

94
95

Figure 8.3 Screen shot of Blocks requiring Planned & Urgent Maintenance
96

Figure 8.4 Screen shot of CFD graph for TGI

Figure8.5 Screen shot CFD of different parameter comparing
two runs

97
 ANALYSIS w.r.t. PARA 607(2) OF IRPWM - 1986
(PRINTED ON - 12/11/2012)
TRC NO. : 7975 RUN DATE : 09-SEP-12
RUN NO. : a DIVISION : CSTM
SECTION : KYN-MMR LINE : UP
KM : 136 To 148
PARAMETER WISE NUMBER OF PEAKS
S KM UNEVENNESS TWIST GAUGE ALIGNMENT
N O FROM TO >10<= >15mm >7.0< >12mm >6mm >5< >10mm
15mm =12mm =10mm
1 136 137 NR NR NR NR NR NR NR
2 137 138 #14 #2 #17 #2 #5 #57 #11
3 138 139 #19 0 #21 #2 #12 #27 #3
4 139 140 0 #1 2 0 #1 3 0
5 140 141 0 0 2 0 0 0 0
6 141 142 0 0 1 0 0 0 0
7 142 143 0 0 1 0 0 0 0
8 143 144 0 0 3 0 0 0 0
9 144 145 0 0 2 0 #1 2 0
10 145 146 3 0 6 #1 #4 #11 0
11 146 147 0 0 1 0 #5 8 0
12 147 148 0 0 4 0 #1 0 0

NOTE :- (1) # Indicates the infringing parameter(s) of the respective

kilometer.
(2) All parameters have been measured in loaded condition.

98
SECTION SUMMARY
UNEVENNESS PEAKS > 10 mm <= 15mm = 2 (km)
UNEVENNESS PEAKS > 15mm = 2 (km)
TWIST PEAKS > 12mm = 3 (km)
TWIST PEAKS > 7mm <= 12mm = 2 (km)
ALIGNMENT PEAKS > 5mm <= 10mm = 3 (km)
ALIGNMENT PEAKS > 10mm = 2 (km)
GAUGE PEAKS > 6mm = 7 (km)

TOTAL KM RECORDED = 11 (km)

KM ON ACCOUNT OF UNEVENNESS = 3 (27.27)%
KM ON ACCOUNT OF TWIST = 3 (27.27)%
KM ON ACCOUNT OF ALIGNMENT = 3 (27.27)%
KM ON ACCOUNT OF GAUGE = 7 (63.64)%
TOTAL (AFTER ACCOUNTING FOR COMMON KM) = 7 (63.64)%

99
100
 Chapter 9

HOW TRC MEASURES THE TRACK PARAMETERS

This chapter may not be directly useful to the field engineers in

understanding the results given by TRC but for the readers with
childlike curiosity this chapter is useful.

9.1 Measurement, Analysis and Reporting – The Past

In the beginning when we started measuring track parameters
manually, the measurements were direct. Gauge was measured
using track gauge by physically placing the lugs of the gauge and
measuring the distance between the two rails at designated points
on inside gauge face. Alignment and unevenness were measured
by stretching a string between 2 points at a distance equal to chord
length e.g. 3.6 or 7.2 or 9.6 m. Cross level was measured by sliding
the spirit level up or down in the gauge cum level instrument. Twist
was then worked out by measuring cross level at two points a given
distance apart e.g. 3.6 m or 4.8 m. Manual measurements are still
done in the same manner.
Three requirements of good track measurement systems are
i) Fast measurements
ii) Quick analysis
iii) Simple reports
Manual method of measurement does not fulfill the above
requirements except for reports being simple.
The first TRC (Amsler Car) had a 3 axle bogie with middle wheel
floating. This middle wheel used to sense the vertical undulations
of the track with respect to reference line joining two outer wheels.
Gauge was measured with the help of sword type contact feelers.
Data acquisition, transmission and recording of track parameters
were all mechanical. Output of the Track Recording Car was in
analogue (graph) form and, therefore, it was not amenable for
computer analysis and all analysis was done manually. Though
the measurement system was fast but the reports were in non-
digital form and therefore the analysis was time consuming. Defect
peaks exceeding A, B, C and D category limits were counted manually

101
 and track categorized into the different categories as explained in
Chapter 2 above. And even the analysis had limitations of revealing
too little about the quality of the track. Recording speed band was
60-80 kmph. That means the results for recording speed beyond
this band were not accurate and thus were ignored.
From this a fourth requirement of good track measurement system
arises
iv) Reliability of results at wide range of recording speeds
From those days to present day the computing and data storage
devices have seen quantum advancements. Similarly motion
sensing devices have seen refinement. When we talk about the
measurement of track, now-a-days, we are talking of measuring
500-700 km of track everyday involving about 10 million
measurements in the present day TRCs. Taking these huge numbers
of measurements either manually or even with old TRC systems is
unthinkable. So we have to have some advanced means of
measurements to achieve this.
9.2 Measurement, analysis and reporting – the present
Without going into details of the developments which have taken
place in the TRC, it would be of more immediate interest of the
readers to know the present system in as much detail (read as little
as is sufficient) so as to give a basic understanding.
Measuring Bogie is an ICF all coil coach with measuring frame fixed
on the four axle boxes of rear bogie as can be seen in the
photographs below.

102
 Figure 9.1 – Photo group showing Measuring Coach and
Measuring Frame
There are four track parameters required to be measured and we
will try to understand the system of measurement of each one of
these in present day TRC – 7965, 7966 and 7967.
9.2.1 Gauge: Gauge is still sensed by sword sensors – sensors as long
as swords. The steel sensors have carbide tips and they touch the
rails constantly during the run. Carbide tips are provided to reduce
the wear of the sensors. The photo below shows the sensor and
the carbide tips.

Figure 9.2– Photo group showing Arrangement for Measuring Gauge

103
 Figure 9.3 Schematic arrangement for measuring gauge
Look at the schematic arrangement shown in figure.9.3 above. The
distance between the hinges ‘H’ where the sword sensors are
supported is fixed and known distance ‘d’. As the TRC moves and
tight gauge is encountered, the sensing portion of one or both arms
marked ‘S’ moves inwards towards the center of track and the upper
ends of the sensor arms marked ‘T’ moves away from the center of
the track. The value of gauge is ‘a1 + d + a2’. Values of a1 and a2 are
worked out measuring b1 and b2.
The upper ends of the sensor arms marked ‘T’ are connected to an
electro-magnetic device called Linear Voltage Differential Transducer
(LVDT) on either side. This device can measure the linear movement
very accurately by measuring the changes in the magnetic field and
the induced voltage differential. Thus the linear movement of points
‘T’ is measured (measurement of distance b1 and b2) and gauge
is calculated. In this case the sensing of the gauge measurement
points is direct whereas the measurement of gauge is indirect. As
explained in chapter 2, a reading of gauge is picked up and stored
in the memory at every 30 cm distance.
9.2.2 Unevenness: Things have moved far ahead from the days of Amsler
when actual measurements of vertical undulations were done by a
3 wheeled bogie as shown below. This is no longer the method for
measuring unevenness. This is just to show the readers how it was
done then. The sketch is self-explanatory. This also explains the
reason for reporting of unevenness results on a chord of 3.6 m.

104
However, systems have been developed in foreign railways using
Laser technology which measures unevenness directly but these
are not technically/commercially matured and have speed limitations
w.r.t. measuring speeds.

Figure 9.4 METHOD for measuring unevenness in OLD TRCs

The method presently used is called the “Inertial system of
Measurement”. This is an indirect method of measurement of
unevenness of rails in the track. To understand this it would be
required to revisit the basic of physics pertaining to distance, time,
speed and acceleration.
Recalling High School Physics, it may be recollected that
differentiating distance with respect to time gives speed and
differentiating speed again with respect to time gives acceleration
at that instant.
If we can measure the acceleration,conversely, we can integrate
acceleration with respect to time to get speed and then integrating
speed again with respect to time we get displacement in that instant.
This is called double integrating acceleration with respect to time.
This is what is done in TRC. Readers would remember,as explained
earlier, that we measure unevenness of left rail and right rail
separately. So we have two accelerometers placed on the coach
floor (inside the TRC coach) – one above the right axle box and other
above the left one. These accelerometers are able to sense the

105
 accelerations due to undulations in the respective rails. Double
integration of these accelerations gives the displacement of the
coach floor – left side and right side respectively. Now this
displacement at the coach floor level can be broken up in two
components –
• undulations of the rails (this can be considered as the vertical
movement of axle box as there is no compressible material
between the top surface of the rail and the axle box)
• compression/expansion of the spring system existing between
the axle box and the coach floor

Figure 9.5 Schematic arrangement of measurement of UNEVENNESS

If we can find out and separate out the second component, we will
get the values of vertical undulations in the respective rails. This is
done by actual measurement of the movement between the coach
floor and the axle box. Again LVDTs are used in vertical direction as
shown in photograph in figure 9.6 & 9.7 below. By this we get the
movement of axle box which reflects nothing but the undulation of
the respective rails. Thus in this way we can plot the absolute profile
of the rail as we are taking these measurements every 30 cm.
This is what is done and therefore it is possible to compute versine
values of unevenness on different chords up to 40 m.

106
 Figure 9.6 Flow Chart for system of measurement of UNEVENNESS

Figure 9.7 Photos showing vertical LVDT arrangements

The values of standard deviation of unevenness are reported for
Left rail and Right rail on short chord of 3.6 m and long chord of
9.6 m.
9.2.3 Alignment : There was no system of measurement of alignment in
Amsler Car. However in the present TRCs the measurement of
alignment is also using “Inertial system of Measurement”. The locus
of center line of the track is computed using an accelerometer
measuring lateral acceleration, mounted on the measuring frame
as shown in the figure 9.8 & 9.9. The distance of the left rail and the

107
 right rail from the center are worked out with the help of LVDT (which
are also used for the gauge measurement). These are also shown
in the figure below.

Figure 9.8 Schematic Arrangement of lateral accelerometer &

LVDTs for measurement of ALIGNMENT

Figure 9.9 Arrangement of lateral accelerometer & LVDTs for

measurement of ALIGNMENT

108
9.2.4 Twist : Twist is worked out by using the absolute vertical profile of
the left rail and right rail and putting an overlapping base of 3.6 m
as well as 4.8 m. This way twist is reported on these two bases.
There is one more method to measure the twist using
inclinometer and gyroscope. Since the value of the gauge is
known, if the inclination (angle) is measured, we can work out
the X-level and hence the twist.
9.2.5 Distance measurement : The details of many of the routes had
been captured with Route Data File Measuring Systems and
exist in the TRC database. The actual distance travelled by the
TRC is measured by measuring the number of full and part
revolutions of one of the wheels of the rear bogie. A tachometer is
fixed to that wheel.
Tachometer consists of a revolving disc with fine slits placed
between a light source on one side and a light sensing diode on
the other side. Each time the light passes through the slit, the
diode receives a signal and it corresponds to a known linear
distance moved. The distance moved will depend on the diameter
of the wheel of the coach and angular distance between the slits.
For picking up reading of the track parameters every 30 cms
(sampling interval), the light passing through the slit sends an
interrupt (signal or order) to the system to save a reading.

Figure 9.10 Schematic arrangement for tachometer

109
 Figure 9.11 Tachometer
9.2.6 Route Data File: The details of the section in linear direction are
collected as a one-time exercise by running a route data
measuring equipment. It not only measures the length of each
kilometer (distance between two kilometer posts, which is
considered a sacrosanct km distance on Indian Railways), but
the location of each of the features like points & crossings, level
crossings, bridge approaches, starting point of curves, SEJs
etc.
Each route is given a unique route code and is downloaded on
the TRC system. You can see this mentioned in the TRC results.
Route code of BCT-ST section is 1786 as can be seen in top
middle portion of the TRC result at Annexure-II. Now when the
TRC runs, this data is recalled from the system and used. The
value of the kilometer from which TRC run commences is fed
and the lengths of all the kilometers following are taken from the
route data file. That means the kilometer punching is no longer
required to be done. Readers who are not familiar with the term
kilometer punching, it is nothing but a manual pressing of switch
connected to TRC system, pressed immediately after each
kilometer is crossed. This is how TRC comes to know that the
kilometer has ended and new one started.

110
 ROUTE DATA FILE
Collection of Route Data and track features
Route Data File Is the Computer ASCII Data File Which
Contains the Detailed Information of Ground Features and its
Location in the Ground. Its use is not only limited to giving
data to TRC car but can also be used to draw a route diagram
of track features. Route data can be created by the methods
given below (i) Trolley mounted TFMS (Track feature
measuring system) now RDPS (Route Data File Preparation
System) (ii) Phase-II Upgraded TRC’s (TGMS – Track geometry
measuring system) (iii) Track Diagrams or (iv) By Extracting
Ground Features from a Run Data File. The system has a
input codes for each of the track features like km post, TP,
P&C, SEJ, LC, Curve start & ends etc. The latest one, RDPS,
has GPS enabled receiver and has a dedicated laptop and
can be run on a trolley up to 30 km/h. Trolley mounted as well
as TRC car based systems use tachometer to measure the
distance.
Synchronization of route data in real time during TRC Run
In spite of using accurately prepared Route Data File,
mismatch in recorded distance of ground features and actual
distance may occur after Running of 20–30 Kms due to
following reasons (i) Wheel Skidding/Slippage (ii) Wheel Wear
(iii) Passing of TRC through loop line instead of main line (iv)
Sinusoidal movement of wheel set. Correction of This
mismatch in actual and recorded distance of ground features
at certain interval is called Synchronization. Synchronization
is done using Automatic Location Detector (ALD) in TRC (fixed
on the cross beam of measuring frame) and Ground
Transponders fixed on track at intervals of 20-30 kilometers
(fixed on sleepers at least at a distance of 200m from the
whole kilometer). Both communicate and the recorded
distance and the distance stored in route data tape is
compared and the mismatch is corrected. But this is in disuse
as the ground transponders of earlier design were very much
theft prone. Now experiments have been done to use RFID
(Radio Frequency Identification) technology for this purpose.

111
 But even for the routes the route data is available in the TRC, it may
happen that errors may creep in due to any of the reasons like TRC
passing from the loop (increasing the distance moved as route
data file is for main line only), variation in the wheel diameter due to
wheel wear,wheel skid/slip and wheel set’s sinusoidal movement.
This error would be significant for distance of 20 to 30 kilometers.
To apply the corrections, TRC had a sensor fixed in its body and
Automatic Location Detectors (ALD) (fixed on two sleepers) in the
track (having its location fed into it) at every 25-30 Km distance
apart. The synchronization of the location is done. Though it is sad
part that ground transponders were prone to theft and were stolen
and these no longer exist in track. So it is advisable to punch
kilometer every 20-25 kms manually to effect manual synchronization
and contain the errors within small stretches of 25-30 kilometers.
9.3 Calibration of TRC
Measuring system has to be calibrated so that it gives the same
measurement for the same entity/defect every time. This is to validate
the measurements. Earlier measurement systems were simple
and easy to understand and it was intuitive and simple to validate.
As the measurements systems became more indirect, the process
going on inside the measurement systems became more opaque,
making it extremely difficult to understand the validation directly. For
this purpose the measuring system is to be calibrated/validated by
comparing it with the known readings. TRC is also calibrated at
periodic intervals. These are two kinds of calibration
(i) Quick calibration – It has to be done daily before starting the
TRC run.
(ii) Detailed calibration – This has to be done
a. Once in a month
b. After replacement of any transducer
c. When quick calibration detects any problem

112
9.3.1 Quick calibration– This is carried out in following steps
a. Bring the TRC outside the platform area to facilitate
measurement of gauge and cross level on a straight and
level track
b. Select calibration option
c. It will display offset counts and input voltage of all transducers,
gauge and cross level
d. Enter recalibrate option (r)
e. Enter measured value of gauge and cross level
f. Message “calibrating” will be flashed till calibration is over
g. Then place 20 mm/10 mm block in front of all LVDTs – two for
the vertical unevenness and two for the alignment
measurement (block to be placed between the sword sensor
and the gauge face of the rail)
h. The change should be 20 mm + 0.2 mm
i. Save settings

Figure 9.11 Calibration for LVDT –

Alignment (left pic.) and Unevenness (right pic.)
9.3.2 Detailed calibration– This may not be required by the field
engineers and is not explained here.

113
 Annexure -I
DEVELOPMENT OF TRACK RECORDING CARS
1.1 Mechanical Track Recording Car
First Mechanical Track Recording Car (Amsler Car) was
introduced on Indian Railways in the year 1963. This car had a
3 axle bogie with middle wheel floating. This middle wheel
was to sense the vertical undulations of the track with respect
to reference line joining two outer wheels. Gauge was
measured with the help of sword type contact feelers. Data
acquisition, transmission and recording of track parameters
were all mechanical. Output of the Track Recording Car was in
analogue form and, therefore, it was not amenable for
computer analysis and all analysis was done manually.
Recording speed band for mechanical TRC was 60 km/h - 80
km/h.
1.2 Electronic Track Recording Car:
To overcome certain limitations of Mechanical Track Recording
Car, the instrumentation system was improved in the year 1975.
In this type of TRC sensing remained mechanical, but
transmission was made electronically. Linear Variable
Differential Transducers (LVDTs) were used to convert
mechanical displacement into electronic voltage resulting in
improvement in accuracies of data transmission. Another
improvement was evaluation of mid-chord versines for the
alignment parameters on 7.2 metre chord from the values of
3.6 metre chord with mathematical manipulation as the mid-
chord versines on 3.6 m chord for alignment did not carry any
meaningful information. The processor for converting
alignment defects from 3.6 m chord to 7.2 chord was developed
by BARC. For the best results recording was to be done between
60 and 80 kmph.
1.3 Track Recording cum Research Car (TRRC):
In the year 1978, the instrumentation for Track Recording-cum-
Research Car was procured from a USA based firm. Measuring

114
bogie of TRRC was also a 3 axle bogie with outer wheels 3.6
m apart & middle wheel 1.8 m apart. Architecture of TRRC was
built around fourth generation on board minicomputer. For the
first time, apart from sensing transmission and recording,
analysing capability was also introduced in the system. Facility
of generating the vertical profile was available and, therefore,
instead of measuring the mid-chord versines on certain fixed
chord, it became possible to measure the defect over user
selectable chord from the absolute profile so generated.
Alignment was measured by chord offset method on 3.6 m
chord and converted to 7.2 m chord mathematically as done in
electronic TRCs. It was also possible to record the vehicle
response like vertical and lateral accelerations, ride indices,
rail wheel forces and axle box accelerations.
It had non-contact capacitive type of sensors for the
measurement of lateral parameters. These contact-less
sensor was not very successful as presence of foreign material
between gauge face and sensing plate often resulted in the
loss of calibration due to change in the capacitance and
frequent calibration of sensors was not possible during the
run. Sensing system was also prone to damage on points
and crossings.
1.4 Micro-processor based Track Recording cars (TRANSYS-
85):
With the development and availability of Micro-processor
Technology and the experience of TRRC it was possible to
further improve the TRCs. This was done in collaboration with
Department of Electronics. This system was developed around
8 bit micro-processor system. Inertial system of track recording
was used to generate absolute vertical & lateral profile. In this
TRC mid-chord versines on one user selectable short/long
chord for vertical and lateral parameters were measured from
generated absolute profile. Statistical analysis of data for the
calculation of standard deviation from mid-chord versines of
various track geometry parameters in a block of 200 metre
was performed on line. In addition to standard deviation, worst
peak information and categorisation of track in A, B, C & D

115
category was also done in real time. In the micro-processor
based track recording cars there were following limitations:-
• It was not possible to record track geometry on two chords
simultaneously which is considered desirable in mixed
traffic scenario.
• There was no facility for data storage for post processing
and data base creation.
• Route data file facility was not available, so punching at
each kilometre was necessary.
1.5 Data Storage System (DSS-5.0):
This system was developed in 1991 and it is popularly known
as phase-I TRC. The existing TRANSYS-85 system was
modified by integrating, one 8086 microprocessor system and
one PC to enhance the capabilities of existing system.
Following additional features in addition to micro-processor
based TRC were incorporated:
• On-line simultaneous processing on two chords. Results
on short chord are vital for freight operations and those on
long chord are more useful for passenger train operation.
• Facilities for storage of processed data for post run analysis.
• Facility of route data file avoiding the need of punching at
each kilometre and other route features.
• Speed band of recording was made wider (50 kmph to 100
kmph) due to which non-recorded length was reduced
considerably.
1.6 Track Geometry Measurement System (TGMS-10):
This was enhancement of the recording accuracy and
processing capabilities of DSS-5.0. The architecture of the
system was redesigned with additional transducers &
hardware in 1992 and designated TGMS-10. Following
additional features were provided in TGMS-10 system. This is
popular known as phase-II upgrade TRC.

116
• Recording of vertical and lateral accelerations as well as
ride indices.
• Facility for raw data storage.
• Facility for synchronisation with ground targets - to avoid
accumulation of error in chainage due to wheel skidding/
slippage,passage of TRC via loop and sinusoidal
movement of wheel set.
• Facility for roll & yaw correction for improving the accuracy
of recording alignment parameters.
• Facility for recording the cross level.
• Speed band was further widened from 36 to 100 kmph with
double sampling of signals for lateral frame accelerometers.
• Facility of analogue as well as digital output.
1.7 Laser based Gauge Measurement System (LGMS):
For recording the gauge and alignment, contact type feelers
were replaced with contact-less sensors consisting of a laser
gun and camera assembly for each rail. Principles of recording
as well as processing remained the same as in earlier TGMS.
Only the contact less sensor had replaced contact feeler and
lateral LVDTs. In case of contact type feelers maximum speed
of recording was restricted to 100 kmph due to feeler dynamics.
With the provision of contact-less sensors it was possible to
record the track parameters up to the recording speed of 160
kmph subjected to max speed of TRC coaches. In addition to
capability of recording at higher speed, improved accuracy of
measurement in lateral track parameter is also expected. This
TRC could not be successful due to validity of trials.
1.8 High Speed Track Recording Car (HSTRC)
A High Speed TRC was commissioned in 2010 on IR. HSTRC
was having “Inertial System” and “Contact-less sensor based
System” for track geometry measurement. Contact-less (laser
based) sensors were used for lateral parameters. This TRC
was also measuring curvature, corrugation defects and

117
misaligned/cupped welds in addition to usual track geometry
and ride parameters.The measurement system of this TRC
has capable of recording upto 200km/h. This was the first TRC
built in IRY shell on IR20 bogie manufactured by Rail Coach
Factory, Kapurthala. This coach was having speed potential
upto 160 km/h.Thus as a whole this TRC was capable of doing
recording upto 160km/h on Indian Railways. However, this TRC
got burnt and destroyed in 2012 due to fire during the run.
1.9 Further area of improvement in the existing TRCs:
• As a fixed frequency cut-off for low pass filter has been
provided in the existing TRC, the cut-off wave length varies
depending upon the speed of recording. A speed
independent low pass filter is desirable.
• At present there is no facility for measuring the small wave
length defect like rail corrugation and misaligned welds
which cause high frequency vibrations and high magnitude
of forces on the rail. In the past, some experiments were
carried out for monitoring the small wave length defect with
the help of recording the axle box acceleration. Output was
in analogue form. Further improvement in this field is
required.
• Recording of curvature at every sampling point.
• Mounting of accelerometer for measurement of lateral
geometry on the coach floor for eliminating the effect of high
frequency vibration.
• Accurate measurement of the defect on the absolute profile
in the full range of wave length of relevance. For this purpose
a facility of changing the cut-off wave length for high pass
filter in certain steps needs to be provided.

118
119 Annexure II – TYPICAL KILOMETER WISE CHART - Actual Result of BCT-VR section – Speed 100 kmph
120 Annexure III – BASIC VALUES OF DIFFERENT PARAMETERS – One time Sheet
 Bibliography
1. Civil & Mechanical Engineering report Volume - 1
2. RDSO report No. TM-1
3. Report of the committee for deciding SUITABLE TRACK INDEX,
Feb’95
4. RDSO report No. TM-11
5. RDSO report No. TM-109
6. RDSO report No. TM-115
7. RDSO report on study done by M/s Transmark of UK for Track
Tolerances
8. 59th meeting of Track Standards Committee, January 1984
9. 66th meeting of Track Standards Committee, June 1993
10. Track Index as a Tool for Speed Certification – Permanent Way
Bulletin, Vol. XXV, No.2, Sep’98 – Arbind Kumar and Hitesh Khanna
11. Track Irregularities and Track Tolerances - Kalicharan
12. Making Track Monitoring more purposeful – Rajeev Bhargava and
Arbind Kumar
13. The little known Phenomenon of Rail Wheel Resonance – M.
Sheshagiri Rao
14. Track Geometry Monitoring – R. S. Varshney
15. Types of track tolerances and their relevance to Indian Railways –
Desh Deepak

121
 Published by :
INDIAN RAILWAY INSTITUTE OF CIVIL ENGINEERING
Designing by : Kalyani Corporation, Pune
Printed by : Sr. Manager, Printing & Stationary C.Rly., Byculla

Price Rs. 70/-

978- 93- 5126- 178- 0