Why Realignment of Curve ?

• The Curve geometry gets disturbed under

passage of traffic as -
• Trains are not moving at equilibrium speed
• Change in Track Stiffness (Ballast condition)
{LC/PC/Br.(?)/Tunnel(?)/etc.}
• Vehicle imperfections
All these Affect SE and alignment
• Additional horizontal forces on the rails due to:
- variations in curvature and SE, incl. transitions
 What is Realignment of Curve ?
• Bringing the curve back to proper alignment - slewing
• Doesn’t necessarily mean restoring to original
alignment, but some desirable geometry, as
Restoration to original geometry may be tedious & time
consuming
May not be feasible due to subsequent developments like
insertion of P&C, OHE mast, Signal posts etc.
• Infinite number of curves are possible between same
set of Tangents
Note: Important is Permissible speed at Proposed
Versines & uniformity of Versines (Variation in
unbalanced forces)
 Realignment of Curve-Types

• Local adjustments – attention to isolated

locations
• Realignments of transitions
 For improving geometry of start & end of curve
 Increasing Transition length-Special case for increasing
speed potential.
• Complete realignment – correction of complete
curve
 Criteria to decide on RoC

• Unsatisfactory running (Foot Plate/Brake Van/

Inspection Carriage/TRC inspections)
• Based on results of curve inspection
Service limits are laid down in IRPWM
 Service tolerances for ROC
Para 524 of New IRPWM)
Limits of station to station variation
Speed range
(mm)
Sectional Speed Below 10 mm (15 mm for speed of 110 km/h) or
160 km/h and 20% of average Versine on circular
upto 110 km/h portion, Whichever is more
Below 110 km/h and 20 mm or 20% of average Versine on
upto 50 km/h circular portion, Whichever is more

40 mm or 20% of average Versine on

Below 50 km/h
circular portion, Whichever is more
Service tolerances for ROC
Para 524 of New IRPWM)
• Local adjustment to be done if the variation is at
isolated few locations.
• If more than 20% station are having versine
variations above the limits prescribed, complete
realignment of curve should be planned within a
month.
So it is the Station to Station variation of versine
which is more important and not the absolute value
of versine.
My suggestion- Realignment must be planned for
each and every curve during tamping irrespective of
the value of Variations in Versines.
 String Lining Method for
Realignment of Curves
(Para-419 of New IRPWM)
Basic Principles
 ROC (String Lining Operation)
Operation-1
•Survey of existing curve by measurement of
versines , on Gauge face of outer rail, and identification of
obligatory points.
• Chord length 20 m
• Versine should be measured 2 to3 stations ahead and 2
to 3 stations beyond last station to ascertain disturbance
of approach tangent track and its rectification, if reqd.

-1
0 +2
-2 +1
ROC (String Lining Operation)
Operation-1 contd..
•In reverse curves, Versine to be transferred to the outer rail at
points where the curvature changes sign. (To use 20/25 mm
special gadget)
•Obligatory points and slew restrictions to be noted.
•In multiple lines - track centre should also be recorded.
•Cant/S.E. at each station to be also recorded.
Note: Any obligatory point is only obligatory in respect of cost
and benefits like if removing obligatory restriction is going to
remove speed restriction then that may be considered for
correction based on cost benefit analysis.
Curve recording pro-forma
RoC (String Lining Operation)

Operation-2
•Determination of the revised alignment and
computation of slews (incl. correct SE)
RoC (String Lining Operation)
Operation-3
•Slewing of curve to revised alignment.
To be carried-out in dry season, unless unavoidable
Note: Curve alignment not be done by eye sight. Proper
computations must be done.
 First Summation of Versine (FSV)

“First summation of versines (Summation

of Versines up to that station) represents
the area of versine diagram (in station
distance units)up to that station
 V0 V1 Vn-3 Vn-2 Vn-1 Vn
V2 V3 V4

If station to station distance is taken as one unit,

Area of each histogram segment = Ordinate at center *1
(i.e., V0, V1, V2, …., Vn-1, Vn etc.)
Total Area of the versine diagram = Sum of areas of each histogram
segment i.e. V0 + V1 + V2 + …. + Vn-1 + Vn
= First Summation of Versines (FSV)
 Second summation of versine (SSV)

Second summation at any station gives the

cumulative figures of First Summation up to
previous station;
and
“Second summation of versines represents the
moment of versine diagram (up to that
station) about that station (in station distance
units)”
 Vn-3 Vn-2 Vn-1 Vn
V0 V1 V2 V3 V4

If station to station distance is taken as unit,

Area of each histogram segment = Ordinate at center
i.e., V0, V1, V2, …., Vn-1, Vn etc.
Lever Arm for each histogram segment wrt last station = n - station number
Moment of versine diagram about station ‘n’
= n* V0 + (n-1)* V1 + (n-2)V2+ ….. + 2* Vn-2+ 1*Vn-1
This is also called second summation of versine values
 First and Second Summations
Station No Versine FSV SSV

0 V0 V0 0

1 V1 V 0+ V 1 V0

2 V2 V 0+ V 1 + V 2 2* V0 + V1

3 V3 V0+ V1+ V2+V3 3* V0+2* V1+ V2

. . . .
. . . .
. . . .
. . .
n-1 Vn-1 .
V0+ V1+ V2 …….+Vn-2+ Vn-1
n Vn V0 + V1 + V2 + …. + Vn-1 + Vn n *V0+(n-1) *V1+ (n-2) *V2+ ….. + 2
*Vn-2+ Vn-1
 1st Property

“The slew in any direction at a station

affects the versines at the adjacent
station by half the amount in the opposite
direction, when the track is not disturbed
at the adjacent stations.”
 Proof

C’
C

B D
b’ Ab = bC = AC/2
b bb’ = CC’/2 (From Similar Δ Abb’ &
ACC’)
Original Versine at B = Bb
A
Modified Versine due to slewing at C ,
Bb’ = Bb – bb’
Bb’ = Bb – CC’/2. Hence proved
Note: The same change will be there at station D
 2nd Property

“The sum total of versines on equal

chord of any number of curves
between the same tangents are
equal.”
 Proof
Consider a curve with
M
only 3 stations with ∆
chord length ‘c’

2α α V1 γ 2γ
β
β K
I
c V2
V0
α γ

L
H
 Proof
• V0 = α*(c/2)
• V1 = β*(c/2)
• V2 = γ*(c/2)

• ΣV = V0 + V1 + V2 = (α+ β+ γ)*(c/2) ------ (1)

• In ∆ IMK, MIK + MKI = Deflection Angle
• i.e. ∆ = (2α+ β) + (2γ+ β) = 2(α+ β+ γ)
• i.e α+ β+ γ= ∆/2 ------------ (2)
• i.e. ΣV = ∆*c/4 (Substituting the value of α+ β+ γ from eqn. 2
in eqn. 1)
So, If station units are constant i.e chord length remains
unaltered and ∆ does not change, the sum of all versines will be
constant
 For a given set of tangents-
Whatever be the actual shape of the curve and /or the starting and end point
of the curve, the summation of versine on same chord remains constant.
 Corollary to 2nd Property

“The chord length being equal, the sum total

of the proposed versines (Vp) should be equal
to the sum total of the existing versines (Ve)
during Curve Re-alignment.”
i.e. ΣVp = ΣVe or Σ Vp – ΣVe = Σ(Vp- Ve) = 0
Or
First Summation of Versine Difference, FSV
(Vp-Ve) = 0
 3rd Property

“The second summation of versine

difference represents half the slew at any
station”
 3’
K 3x2V0

L 2x2V1

M 2xV2
2’ 2x2V0 N
G
3
H 2V1
I
V2
J
2
1’ D
E
2xV0
F V1
1

The offset from straight for the station no. n:

B
0
C
V0 2*n*V0+2*(n-1)V1+2*(n-2)V2+ … 2*2 *Vn-2+ 2*Vn-1

-1
A
 Second summation application
The offset from straight for the station no. n:
•2*n*V0+2*(n-1)V1+2*(n-2)V2+ … 2*2 *Vn-2+ 2*Vn-1
•2*(n*V0+(n-1)V1+(n-2)V2+ … 2 *Vn-2+ 1*Vn-1 )
•2*(Second summation of versine till station ‘n’)
It means twice the second summation of versines of a
curve upto a station represents the offset of curve from
tangent to that point.
From Twice the difference of second summations of the
existing and proposed versines, we get a slew at a station
i.e
Slew = 2* SS of Vp – 2* SS of Ve
= 2* SS of (Vp- Ve)
 Second summation application contd..

2SSV of
Vp

Slew = 2SSVp - 2SSVe

2 SSV = 2* SS (Vp-Ve)
of Ve Or SS (Vp-Ve) = Slew/2

Sign convention → If slew = +ve , it is towards inside of curve

If slew =-ve , it is towards outside of curve
 4th Property

• Second summation of the difference of the

Proposed and Existing Versines at first and
last station should be zero
- Not sacrosanct but must be kept to avoid disturbance of
the area outside the dealt portion
 Steps in String Lining
• Survey existing versines (Ve) after gauging
• Find sum of existing versines and get an idea of the
average versines {=Σ Ve/(C+L)} in circular portion
• Propose new versines (Vp) for the curve according to
the following principles:
• Sum of existing versines(Ve) = Sum of proposed versines
(Vp)
• Uniform rate of change of versine in transition portion
• Uniform Proposed versine {≈Σ Ve/(C+L)} over circular
portion
 Steps in String Lining (Contd.)
• Workout Versine Difference (Vp-Ve) for each
station
• Workout First Summation of Versine Difference
- FS at last station shall be zero (If the values proposed are
correct) as per Corollary to 2nd Property
• Workout Second Summation of Versine Difference
• Value at first and last station shall be zero (If the values
proposed are actually desired ones)
Note: Normally some residual slew is left either inward
or outward at the end i.e. second summation at last
station is not zero as the curve is disturbed and/or due
to rounding off of proposed versines.
Station Existing 8 32 3
number versines
9 28 3
in mm or
20 M 10 36 3
chord 11 34 3
(1) (2) (3) 12 32 3
0 2 2 13 34 3
14 36 3
1 0 8
15 24 3
2 14 16
16 24 2
3 28 24
17 28 1
4 30 32 18 0
5 36 32 19 6
6 36 32 20 0
7 24 32 484
Station No. Ve Vp
(1) (2) (3)
0 2 2
1 0 8
2 14 16 vp = Σve /(L+C)
3 28 24
4 30 32
5 36 32
6 36 32
vp = 484/15
7 24 32 = 32.2
8 32 32 say 32
9 28 32
10 36 32
Where L=5,
11 34 32
12 32 32 C=10 stn
13 34 32 units
14 36 32
15 24 32
16 24 24
17 28 16
Σ Vp
18 0 8 should be
19 6 2 484
20 0 0
Station Existing Proposed Versine 1st 2nd
number versines versine in difference summation summation
in mm or mm (3) - (2) of versine of versine
20 M difference difference
chord or half
throw in
mm

(1) (2) (3) (4) (5) (6)

0 2 2 0 0 0
1 0 8 +8 +8 0
2 14 16 +2 +10 +8
3 28 24 -4 +6 +18
4 30 32 +2 +8 +24
5 36 32 -4 +4 +32
6 36 32 -4 0 +36
7 24 32 +8 +8 +36
8 32 32 0 +8 +44
9 28 32 +4 +12 +52
10 36 32 -4 +8 +64
11 34 32 -2 +6 +72
12 32 32 0 +6 +78
13 34 32 -2 +4 +84
14 36 32 -4 0 +88
15 24 32 +8 +8 +88
16 24 24 0 +8 +96
17 28 16 -12 -4 +104
18 0 8 +8 +4 +100
19 6 2 -4 0 +104
20 0 0 0 0 +104
 Steps in String Lining (Contd.)

• If SS at last station is non-zero, it means that

The moment of the Existing and proposed versine
diagrams about the last station are not equal
• Or The C. G. of both the versine diagrams are not
coinciding.
• Or
The existing and proposed curve are not merging at same
location at that end since Slew is not zero.
SS(Vp-Ve) and Correcting couple ?
C.G. of

C.G. of

If C.G. of Vp is to left of Ve, +ve

moment about last station

+ive
-ive
 Steps in String Lining (Contd.)
• This residual slew is to be eliminated by applying
reverse moment or correcting couple such that
Resultant of SS of (Vp-Ve) and that of CC at last
station becomes zero.
 Correcting couple (CC)

• Correcting couple – Small correction in versine diagram

- by reducing versine at one location and
increasing the versine by equal amount at other station
• Applying correcting couple means the transfer of area
of versine diagram so that C.G of existing and proposed
versine diagram matches.
 Correcting couple(CC) contd..

• Location of Application of CC
At max. distance apart for maximum effect.
• Principles for choosing CC
Values of CC - Equal and opposite
Sign of correcting couple
- Opposite to the residual second summation at the last
summation
{meaning if residual second summation is + 104, then it will be –ve
at initial station of curve (giving larger moment) and +ve at last
end of curve (giving lesser moment about curve end)}
Net moment of CC = -Vcc * x +Vcc*y (x, y are distances from curve
end, assuming x > y)
= -Vcc *(x-y)
= value of moment * distance between them.
 Correcting couple(CC) contd..

• Principles for choosing CC contd..

As small as possible – as it disturbs ideal trapezoidal shape of
versine diagram
At max. distance apart for more effect with least disturbance to
trapezoidal shape
A no. of small correcting couples distributed throughout the
length, if one is not sufficient till residual SS at last station
becomes zero.
 Correcting couple (CC)
• Why Correcting couple ?
We can do without applying correcting couple and try with
some other assumed versine value.
• Why correcting couple has to be equal and opposite?
To keep sum of versine constant
Steps in String Lining (Contd.)

• More than one correction couple may be required till we

get its summation equal to second summation of versine
difference .
• Workout FS and SS for CC
• FS to be zero at last station and SS of CC at last station
should be equal and opposite of SS of (Vp-Ve)

With Excel, we can have more flexibility to decide changes

required in versine
Station Existing Proposed Versine Connecting couple
Correcting Couple
number versines versine in difference 1st 2nd Connecting 1st 2nd
in mm or mm (3) - (2) Correcting
summation summation versine in summation summation
20 M of versine of versine mm of of
chord difference difference correcting correcting
or half versine versine
throw in
mm
(1) (2) (3) (4) (5) (6) (7) 98) (9)
0 2 2 0 0 0 -1 -1
1 0 8 +8 +8 0 -1 -2 -1
2 14 16 +2 +10 +8 -1 -3 -3
3 28 24 -4 +6 +18 -1 -4 -6
4 30 32 +2 +8 +24 -1 -5 -10
5 36 32 -4 +4 +32 -1 -6 -15
6 36 32 -4 0 +36 -1 -7 -21
7 24 32 +8 +8 +36 -1 -8 -28
8 32 32 0 +8 +44 -8 -36
9 28 32 +4 +12 +52 -8 -44
10 36 32 -4 +8 +64 -8 -52
11 34 32 -2 +6 +72 -8 -60
12 32 32 0 +6 +78 -8 -68
13 34 32 -2 +4 +84 +1 -7 -76
14 36 32 -4 0 +88 +1 -6 -83
15 24 32 +8 +8 +88 +1 -5 -89
16 24 24 0 +8 +96 +1 -4 -94
17 28 16 -12 -4 +104 +1 -3 -98
18 0 8 +8 +4 +100 +1 -2 -101
19 6 2 -4 0 +104 +1 -1 -103
20 0 0 0 0 +104 +1 0 -104
Steps in String Lining (Contd.)

• Workout Resultant Versines (=Vp+ CC)

• Add the SS for original versine difference and the SS for
the correcting couple
• Workout Resultant Slew (These slews are to be actually
applied in field) = 2* {SS of (Vp – Ve) + SS of CC}
Station Existing Proposed Versine Correcting couple
Connecting
number versines versine in difference 1st 2nd Connecting 1st 2nd Resultant Resultant Resultant
Correcting
in mm or mm (3) - (2) summation summation versine in summation summation half slew full slew versine in
20 M of versine of versine mm of of (9) + (6) inmm mm
chord difference difference correcting correcting (3) + (7)
or half versine versine
throw in
mm
(1) (2) (3) (4) (5) (6) (7) 98) (9) (10) (11) (12)
0 2 2 0 0 0 -1 -1 0 0 1
1 0 8 +8 +8 0 -1 -2 -1 -1 -2 7
2 14 16 +2 +10 +8 -1 -3 -3 +5 +10 15
3 28 24 -4 +6 +18 -1 -4 -6 +12 +24 23
4 30 32 +2 +8 +24 -1 -5 -10 +14 +28 31
5 36 32 -4 +4 +32 -1 -6 -15 +17 +34 31
6 36 32 -4 0 +36 -1 -7 -21 +15 +30 31
7 24 32 +8 +8 +36 -1 -8 -28 +8 +16 31
8 32 32 0 +8 +44 -8 -36 +8 +16 32
9 28 32 +4 +12 +52 -8 -44 +8 +16 32
10 36 32 -4 +8 +64 -8 -52 +12 +24 32
11 34 32 -2 +6 +72 -8 -60 +12 +24 32
12 32 32 0 +6 +78 -8 -68 +10 +20 32
13 34 32 -2 +4 +84 +1 -7 -76 +8 +16 33
14 36 32 -4 0 +88 +1 -6 -83 +5 +10 33
15 24 32 +8 +8 +88 +1 -5 -89 -1 -2 33
16 24 24 0 +8 +96 +1 -4 -94 +2 +4 25
17 28 16 -12 -4 +104 +1 -3 -98 +6 +12 17
18 0 8 +8 +4 +100 +1 -2 -101 -1 -2 9
19 6 2 -4 0 +104 +1 -1 -103 +1 +2 3
20 0 0 0 0 +104 +1 0 -104 0 0 1
 Shortcomings in Method
• Difficult to decide proposed versines, especially when the stations
are more (suitable for small curves only)
• Curve often get disturbed so much that slewing to original
alignment may require slews more than practicable limits (50 -
100mm) –
start & end points may require alterations for practicable slews
with good riding quality
Repeated iterations for residual slew at either end with
practicable slews
• There is no way to know, if the length of existing curve is ok
Curve may increase or decrease in length during service due to
elongation/shortening of rail
affects assumed proposed versines
• Correct beginning and/or End of curve is not known
 SITUATIONS POSSIBLE
1. Improper Choice of beginning of curve:

2. Improper choice of versine:

 Optimization method

• Optimization method –
for establishing correct beginning & end of curve &
versines
Fixing the centre of gravity of versine Diagram of the
curve and ensuring that it remains at the same place for
proposed versines also.
 Realignment Programs available (at IRICEN website)

Program by Sh M. S. Ekbote
Has modules for ROC of simple curves by
 Pure trapezoidal solution
 Realignment in segments
 Averaging method
 Limited maximum slews
Incorporates obligatory points
ROC method By Mr Abu Hasan Has modules for ROC of
 Reverse curves
 Only transitions
 Vertical curves
Software for Curve Design, Existing Speed, Increasing Speed,
Re-alignment & Track Machine Data- By Shri Anil Choudhary, Sr.
Prof. /Track Machine, IRICEN
Software for Curve Re-alignment – Optimum Solution – by Shri
R.K.Shekhawat, Sr. Prof.(Projects), IRICEN
 Solving real life problem
Q. The curve given in previous example has the following
particulars –
Transition length, L= 50m (Both sides), Circular length = 100m
Existing Cant = 60mm, Radius of curve = 1560m
Min. booked speed of Goods train = 65 kmph
Max. sectional speed = 130 kmph.
a) Calculate the speed potential of existing curve.
b) Realign the curve as per existing particulars.
c) Redesign the curve for the sectional speed
d) Realign the curve accordingly
Thank you