Guidelines to

Best Practices
For Heavy Haul
Railway Operations:
Management of the Wheel
and
Rail Interface

First Editor: John Leeper

Second Editor: Roy Allen

Published by Simmons-Boardman Books, Inc.

i
This book is the third in a series: Guidelines to Best Practices for Heavy Haul
Railway Operations

These Guidelines were commissioned by the International Heavy Haul Associa-

tion (IHHA) and its Board of Directors as input to the decision-making processes
of heavy haul railways. They represent the best efforts of IHHA and the interna-
tional team of leaders in the field of heavy haul operations and research. Their
collective experience and knowledge formed the basis of the book. However,
there are circumstances in which the best practices may differ from those discussed
in the Guidelines. Therefore, these Guidelines are neither mandatory nor do
they describe exclusive methods to achieve optimum rail/wheel performance.

Copyright © 2015 International Heavy Haul Association

All rights reserved.

Reproduction or translation of any part of this work without the permission of

the copyright owner is unlawful. Requests for permission or further information
should be addressed to:

International Heavy Haul Association

2808 Forest Hills Court
Virginia Beach, Virginia 23454 USA

ISBN 978-0-911382-63-1

Library of Congress Control No.: 2015938152

Printed in the United States of America

Book and design layout by Simmons-Boardman Books, Inc.,

1809 Capitol Ave., Omaha, NE 68102

ii
The following chapter is excerpted from:

Guidelines to Best Practices For Heavy Haul Railway Operations:

Management of the Wheel and Rail Interface
Chapter 2: The Fundamentals of Vehicle / Track Interaction

2.1 Introduction
The transverse shape and longitudinal running profile of wheel and rail are
essential to the effective guidance of a rail vehicle on the track. Wheel/rail
contact will change these shapes and profiles over time through wear,
deformation, and (in the limit) failure of the material on the running surfaces.
A change in shape can reduce the effectiveness of vehicle guidance. Material
failure of the running surface compromises the structural integrity of the wheel
and/or rail. Compromised vehicle guidance and failure of the running surface
will accelerate vehicle and track degradation.
The task of vehicle and track designers is to provide cost-effective products
that reduce wheel and rail forces, stresses, deformations, and wear, retaining
desirable wheel and rail shapes as long as economically possible.
The task of railroad operators is to choose appropriate designs for their
operating circumstances and maintain vehicles and track, and in particular the
contact interface, within limits to provide for safe and cost-effective operations.
Operations can differ significantly from one railroad to another as they can
have very different parameters such as track topology (curvature and grade),
axle and train loads, gage, climate, geology, operations, and past operational his-
tory. Any one of these parameters can significantly influence the many choices
of wheel/rail interface design or maintenance practices used. Consequently, no
single “recipe” for success exists as there is a large list of possible “ingredients.”
An informed understanding of the basic mechanisms of wheel/rail and vehicle/
track interaction can assist in choosing products and strategies and avoiding
potential conflicts. The object of this chapter is to describe these basic mecha-
nisms and their impact on wheel/rail interaction. This description will be qualita-
tive and should help the reader to make informed decisions on product choice
or maintenance intervention. Quantitative answers are best left to experts in the
many different fields involved.
Note that the IHHA published a handbook on wheel/rail interaction in 2001.2.1
In the ensuing 14 years, the industry knowledge base has broadened and more
tonne-kilometers (ton-miles) are being operated at elevated loads. Wagon and
bogie design and track maintenance practices are evolving to suit.
The science of contact mechanics under heavy haul operations is being
increasingly better understood, particularly the role of steering tractions in
the development of rolling contact fatigue. This has led to the development
and implementation of improved steering vehicles and bogies and the increas-
ing application of contact-conditioning technologies such as top-of-rail friction
control. The track has been strengthened and wheel profiles and grinding
practices implemented to counter rail rollover. Wheel steels are being im-
proved and wheel machining practices updated to counter the incidence of tread
shelling and wheel failures such as vertical split rims. Rail materials and
cleanliness are continually being upgraded.

2-1
 This chapter has been revised to reflect these changes. Heavy haul
development continues and the reader should expect further developments in
the coming years, particularly in the science of contact mechanics (reference
chapter 3).
Some of the typical questions that may be answered in this chapter are:
• How does a wheelset/rail vehicle steer on the track?
n What is the importance of wheel and rail profiles?
n What is conicity?
• What is vehicle hunting?
• Is there a compromise that must be made between vehicle curving and
hunting?
• What significance do pitch, bounce, and rock and roll modes have in
vehicle/track interaction?
• What generic forces are generated between the wheel and the rail and
what damage might these forces do to the track? (reference also
chapter 3)
• Considering the basic mechanisms of wheel/rail and vehicle track
interaction:
n What are the main system parameters I must consider when designing
a heavy haul operation?
n What are the important maintenance parameters I can control?

2.2 Railway Wheelset and Track

The railway wheelset is traditionally comprised of two wheels that are fixed
rigidly to a common axle (see figure 2-1). Wheelsets with independently rotating
wheels are being used to a limited degree on certain passenger rail vehicles, but
not in heavy haul applications.
The rolling surfaces of the wheels, the wheel treads, are cut to a cone angle, γ.
Typically, more complex profiles termed “hollow,” “worn,” or “profiled” treads
are used. These profiles generally have a nonlinear conicity characteristic.2.2

Figure 2-1: The Railway Wheelset

2-2
 The track comprises two rails laid on sleepers at a particular gage, as
figure 2-2 shows. The rails are laid at an angle, β, to the sleeper to generally
match the angle, γ, of the wheelset profile. This centralizes contact on the center
of the head of the rail on tangent track and assists in stabilizing the rail against
rollover as the normal reaction to the contact as the wheel generally passes
through the base of the rail.

Figure 2-2: Cross Section of the Railway Track

Generally, the distance between the centers of the rail heads (the gage plus
the width of one rail head) equals 2l as figure 2-2 indicates. This is also, gener-
ally, the distance between the so-called “tape lines” on the wheelset (see also
figure 2-1). The tape line indicates the lateral position on the wheel tread around
which the circumference of the wheel is measured using a tape. The wheel diam-
eter is inferred from the circumference. The difference in diameter between the
two wheels on the same axle is generally of the order of 1 mm (0.04 in.).
A layer of ballast supports the sleeper. The ballast permits alignment
adjustment, as well as vertical, lateral, and longitudinal stabilization of the track.
It further provides some vertical resilience to passing trains. The structure of the
ballast also provides protection to the track substructure by spreading the load
and by dissipating vibration energy. The voids in the ballast permit drainage and
a degree of accumulation of fine material, without any significant change to the
alignment or resilience of the track.
Railway track is generally “canted” or superelevated in curves to counter
centripetal forces without appreciably transferring wheel loads between out-
er and inner rails as the vehicle negotiates the curves at a higher speed, as
described in section 2.8.1. In the limit, superelevation helps prevent overturn-
ing of the vehicle. However, inappropriate matching of superelevation to vehicle
speed can adversely influence the curving performance of the vehicle and, in
turn, the wear and stresses in rail and wheel.
The portion of track between tangent and curved track is termed a spiral
transition curve, and the vehicle experiences this as track twisted about a
longitudinal axis. This implies that the contact patches on the different wheels
may not be in the same plane. This would lead to a loss of normal load between

2-3
some wheels and the rail if excessively stiff vertical suspension designs are used.
Furthermore, the running surface of the rail is discontinuous at nonwelded rail
joints and certain types of crossings at turnouts. This can cause impact loads on
the wheel and the rail and, momentarily, result in a shift in the wheel/rail contact
position on the wheel profile. This may affect wheelset guidance.
Steel-on-steel contact produces a uniquely low rolling resistance for railway
vehicles. The geometry of the wheelset, described mathematically as a di-cone
(two cones placed back to back having a cone angle, 2γ), imparts on the wheelset
unique properties of self-guidance (i.e., self-centering on tangent track as figure
2-3 shows and the ability to negotiate curves as shown in figure 2-4).
The di-cone design also provides the railway wheelset with the ability to
accommodate diameter inaccuracies between the two wheels by displacing lat-
erally on tangent track to compensate for the diameter difference. The flange
on the wheel tread is used where track discontinuities or track geometry
irregularities are so severe or the vehicle suspension is so inadequate that the
properties of self-guidance of the wheelset are insufficient for guidance without
flange contact.

Figure 2-3: Self-Centering Motion on Tangent Track

Figure 2-4: The Generation of a Radius Differential in a Curve

2-4
2.3 Kinematics of a Railway Wheelset

2.3.1 Introduction
The dynamic interaction between the vehicle and the track, and thus between
the wheel and the rail, has a significant influence on the running safety and ride
quality of railway vehicles as well as on the fatigue and wear of the wheel and the
rail surface. It is thus essential to have a good understanding of the impact of,
for example, heavier axle loads on critical system performance issues and their
impact on safety and system degradation.
The wheel tread is generally conical or has a particular curvature in the
transverse direction (described in section 2.2). Figure 2-5 shows the wheel tread
and other typical features of a wheel profile.

Figure 2-5: Typical Wheel Profile

Similarly, rails also have a curved profile in the transverse direction
(figure 2-6).

Figure 2-6: Typical Rail Profile

These geometric features contribute to the behavior of the railway vehicle as
a dynamic system.

2-5
2.3.2 Rolling Radius Differential
The difference between the rolling radii on a wheelset is known as the “rolling
radius difference” and is an important parameter in analyzing the behavior of the
railway wheelset.
Figure 2-7 shows the rolling radius difference, Δr, as defined by:
Δr = γ y (2.1)
where y is the lateral displacement of the wheelset from the center position of
the track and γ is the conicity of the wheel tread. For conical wheels, the conicity
is equal to the contact angle of the wheel profile.

3 N P₀,max =
(Note that the figure includes Redtenbacher’s formula2.3P₀(ξ,η)
) = 2 Ae √ 1 - ( aξ )² - ( ηb )² ,
Figure 2-7: Rolling of a Coned Wheelset on a Curve
(32) =
τ₁,max = 0.3 ∙ p₀ at z = 0.78a.
2.3.3 Conicity
For wheelsets with conical wheels, the conicity is derived from the mean slope
of the rolling radius versus lateral wheelset displacement and
Rw δis defined as: r0 l
Δr γe = y =
(2.2) γ = y R w - RR Rc γ
For circular wheel/rail contact geometry (figure 2-8 is an example) a so-called
“equivalent conicity” is derived from equation (2.2) in terms of the profile radii:
f E
kb² = 11 kp l²
f22
Δr = r0 R + l - 1
R-l [ ]

Δ r γ y R R δ2l λ π w e

Figure 2-8: Profiled Wheel and Rail

2-6
 (32) = (1 tan(δ)
+ μ * tan(δ))
-μ
τ₁,max = 0.3 ∙ p₀ at z = 0.78a.

lr
(2.3) Rw δ r l λ = 2π 0
γe = y = 0 γ
γ = Δr R - R y w R Rc γ

Where Rw is the wheel profile radius, RR is the rail profile radius, and δ is the
angle between the plane f11of contact and track level.
2.1, 2.2
Effective Conicity (λ) = R

reduces to:
kb² =
When Rw is set to infinity
f22
k l²
and
p Rw
is smallΔr =
(ar R+l -1
[ ]
0typical rail profile), equation (2.3)
R-l
2

γe = δ (2.4)
or the angle of the wheel profile. γ = rolling radius
2y

2.3.4 Curving Action - Pure Rolling

Coning of the wheel tread was well established by 1821. Not only was the use
of tapered cast-iron wheels part of normal foundry practice, but coning was also
used to reduce the rubbing of the flange on the rail and to ease the motion of the
vehicle in curves.
In 1855, Redtenbacher2.3 provided Δ r the γ y first 2l λ π
R R δtheoretical analysisw
of the
e curving
R
behavior of a single, unconstrained wheelset. Figure 2-7 shows that there is a
simple geometric relationship
3 N between the P₀,max outward - NA
= 2 movement
3
of the wheel
P₀(ξ,η) = Ae √ 1 - ( aξ )² - ( ηb )² , e
y, the radius of the2 curve R, the wheel radius r0, the distance between the
contact points 2l, and the conicity γ of the wheels
τ₁,max = 0.3 ∙ p₀ at z = 0.78a.
( tan(δ) - μ
(32) =in order )
to sustain so-called
1 + μ * tan(δ)
“pure rolling.”
The lateral displacement of the wheelset from the center of the curve for “pure
rolling” is given by: lr v
Rw δ r0 l λ = 2π 0 f = 0
γ = y
γ = (2.5) γ
Δr = λ
y
e R w - RR Rc γ
The application of Redtenbacher’s formula shows that a wheelset will only be
able to move outward to achieve pure rolling if the radius of curvature, or the
f Effective Conicity (λ) = RRD
flange
kb² = 11waykp l²clearance, orΔrthe= r0 conicity
[
Thef22term “pure rolling” is Ra - kinematic
l ]
R + l - 1is sufficiently large.
term describing the motion of an
2y

unconstrained massless wheelset not subject to any external forces. The “pure
rolling” position thus describes a “zero force” state of the wheelset. Deviations
rolling radius differential
from this position result in the generation of forces between γ =wheel and2yrail.
On curved track, the outer wheel of a wheelset has a greater distance to travel
than the inner wheel. To compensate for this, the wheelset thus moves laterally
relative to the track so that the larger wheel radius on the inner side of the wheel
runs on the outer rail of the curve. The inner wheel uses the outer side of its
tread to reduce the traveled distance during the passage through the curve. This
causes each wheel to run a different distance per revolution of the axle. Hence,
the wheelset alignsΔ r γitself
y R Rradially
δ2l λ π in curves. e
w R

2-7
2.3.5 Self-Centering Action
The geometry of the wheelset allows it to negotiate curves due to the radius
differential generated between the wheels on the common axle, as well as the
unique property of self-guidance [i.e., self-centering on tangent track (figure
2-9)].

Figure 2-9: Self-Centering Kinematic Oscillation of a Wheelset on Tangent Track

George Stephenson in his “Observations on Edge and Tram Railways” stated:2.4
“It must be understood the form of edge railway wheels are conical that is the outer
is rather less than the inner diameter about 3/16 of an inch. Then from a small
irregularity of the railway the wheels may be thrown a little to the right or a little
to the left, when the former happens the right wheel will expose a larger and the left
one a smaller diameter to the bearing surface of the rail which will cause the latter
to lose ground of the former but at the same time in moving forward it gradually
exposes a greater diameter to the rail while the right one on the contrary is gradually
exposing a lesser which will cause it to lose ground of the left one but will regain it
on its progress as has been described alternately gaining and losing ground of each
other which will cause the wheels to proceed in an oscillatory but easy motion on
the rails.”
This self-centering property has the following benefits:
• In a quasi-static mode, it can compensate for tolerances in the diameters
of the two wheels on the wheelset by moving (quasi-statically) laterally on
the track.
• In a kinematic sense, the wheelset can centralize itself on the track after
being laterally displaced or encountering a lateral track irregularity. Thus,
if a wheelset is rolling along the track and is displaced slightly to one side,
the wheel on one side is running on a larger radius and the wheel on the
other side is running on a smaller radius. Because the wheels are
mounted on a common axle, the wheel with the larger radius will move
forward faster than the other. Thus, if pure rolling is maintained, the
wheelset moves back toward the center of the track.

2.3.6 Wheelset Hunting

The wheelset, however, generally overshoots the center of the track or
“understeers.” This causes a sustained kinematic oscillation called “hunting.” A
free wheelset with a finite mass experiences this type of instability at all forward
speeds.

2-8
 Hunting is an oscillation, usually unwanted, about an equilibrium position.
The expression came into use in the 19th century and describes how a system
“hunts” for equilibrium. A classical hunting oscillation is the yawing motion of a
railway vehicle caused by the coning action on which the directional stability of
a3 railway
N depends. P₀,max = 32 - NA
P₀(ξ,η) = 2 Ae √ 1 - ( aξ )² - ( ηb )² , e
In 1883, Klingel gave3 the firsttan(δ)
2.5
mathematical analysis of the kinematic
oscillation of an
τ₁,max = 0.3 ∙ p₀ at z = 0.78a.
unconstrained (
(2) = wheelset
1 + μ *
) -μ
and
tan(δ) derived the relationship between
the wavelength λ and the wheelset conicity γ, wheel radius r0, and the lateral
distance between 3contact N points between wheels and rails 2l as:
P₀,max = -
√ 1 - ( aξ )² - ( ηb )² , e 2 A
lr v
R
wδ (2.6)
l -μ λ = 2π 0 f = 0
()=( )
r
tan(δ)
γe = 3 y = 0 γ λ
R w - RR
₀ at z = 0.78a.
2 Rcμγ* tan(δ)
1+
According to Klingel’s formula, the frequency of the kinematic oscillation will
increase with increasing speed. The frequency of the kinematic oscillation is
defined as: Effective Conicity (λ) = RRD
p l²

RR
y = 0 [
Δrr =l r0 R + l - 1
R - l ]
(2.7) λ = 2π
lr0
γ
f =
v0
λ
2y
Rc γ
Lateral oscillations caused by coning were experienced from the early days
of the railways. One solution to the oscillation problem that has been proposed
γ = rolling radius differential
from time to time, even
Effective down
Conicity to RRD times, was
modern 2y to fit wheels with cylindrical
= r0
[ R+l -1
] (λ) =
2y
R - l treads. However, in this case, if the wheels are rigidly mounted on the axle, very
slight errors in parallelism would induce large lateral displacements that would
be limited by flange contact. Thus, a wheelset with cylindrical treads tends to
run in continuous flange rolling radius
γ = contact. It isdifferential
probably for this reason and the resulting
accelerated wear that cylindrical wheel 2y treads have never been used successfully
in heavy haul service.
Δ r γ y R R δ2l λ π w e
2.4 Quasi-Static Forces Acting on a Railway Wheelset R

√
2.4.1 Introduction
R δ2l λ π In this section,e
the forces acting in the wheel/rail contact are introduced.
w
These forces, to a great
R
extent, determine the running behavior of the vehicle.
√
They are, however, equally important to understand the damage mechanisms on
wheels and rails introduced later in chapter 3. For the sake of simplicity, only the
quasi-static situation is described where dynamic oscillations of the vehicles are
not taken into account.

2.4.2 Creep and Creep Forces

Forces imposed on a wheelset as a result of constraints and/or forces applied
at the journals cause the wheelset to deviate from a pure rolling position on the
track. Partial sliding motions take place across the contact patch resulting in
relative slip called creep or creepage. Creep results in the generation of forces
across the contact patch to oppose this dynamic action.
The creep forces in the wheel/rail contact depend on the creep, the normal
load in the contact, the friction coefficient, and the contact geometry. The
relationship is nonlinear. Figure 2-10 shows a typical relationship between creep
and creep force.

2-9
 Figure 2-10: Creep Force as Function of Creep
The following interesting observations can be made:
• When creep is zero, the creep force is zero. In other words, to develop a
creep force, creepage in the wheel/rail contact is needed. This means
for example, to develop a traction or braking force, longitudinal creep in
the wheel/rail contact is needed.
• For small creepages, the relationship is linear. The creep force, however,
saturates at the value μN (i.e., this is the highest value of creep force
possible to transfer in the contact).
• Given knowledge of the creep relationship (example figure 2-10), if the
position of the wheelset on the track is known, the creep may be
calculated together with the creep force.
The creepages and resulting forces acting across the contact patch can be di-
vided into three components:
• Longitudinal creep and resulting force
• Lateral creep and resulting force
• Rotational creep around the axis normal to the contact patch, which is
called spin or spin creep and results in a reacting moment and coupled
lateral force
The direction and magnitude of these creepages and creep forces are explained
in the following paragraphs.

2.4.2.1 Longitudinal Creep and Creep Force

Longitudinal creepages arise if the forward speed of the wheel does not
correspond to the product of rolling radius and rotational speed. Longitudinal
creep exists for example during acceleration and braking when there is over-
or under-speed in the wheels. Further longitudinal creepage is present when a
wheelset is laterally displaced from the pure rolling position. On straight track,
the pure rolling position is the centered position. Due to the rolling radius
difference, the wheelset would like to yaw against the track to the preferred
final state marked with a dash-dotted line in figure 2-11. It is forced, however,

2-10
into a position at right angles to the track center line by forces FJ applied by the
vehicle suspension to reach the final actual state shown in figure 2-11. This is
only possible if the wheel rolling on the larger radius slips backward and the
smaller wheel slips forward. This slip induces forces FS.

Figure 2-11: Longitudinal Creep and Creep

Forces on Tangent Track
Similarly if the wheelset is displaced toward the high rail from the pure rolling
position (where the rolling radius difference matches the curve radius), creep
forces FS are generated (figure 2-12).

Figure 2-12: Longitudinal Creep on

Curved Track

2-11
2.4.2.2 Lateral Creep
Lateral creepage can be explained in a similar way. For a wheelset placed at an
angle of attack α, the angle between the wheelset axle and a line perpendicular
to the track center line is shown in figure 2-13. A pure rolling wheelset would
roll off the track to the preferred final state indicated by the chain-dotted lines in
figure 2-13. As the wheelset is constrained by forces FJ acting at the journals (as
shown in figure 2-12) and by possible flange forces, it has to stay on the track.
This is only possible with a slip motion, directed to the outside of the curve, if
the wheelset moves forward with velocity V. The amount of lateral creepage is
proportional to the attack angle of the wheelset against the track.

2.4.2.3 Spin Creep

The magnitude of the spin creep generally depends on the contact angle γ
(figure 2-14). Imagine removing the connection between the wheels. In that
case, coned wheels would roll in a circle instead of following the track. To follow
the track, a constant rotational slip motion between wheel and rail is necessary.
This means there is spin creepage also for a wheelset running in the centered
position on straight track as long as the contact angle is not zero.

2.4.2.4 Significance of Creep Forces on Vehicle Behavior

Generally, the longitudinal and lateral creep forces play the most significant
role in vehicle behavior. If controlled correctly, they provide effective guidance
to the vehicle. If not, they can be the source of much track and wheel damage.
Spin creep generally does not play a significant role in vehicle behavior.

Figure 2-13: Lateral Creepage on Tangent Track

For more information on creep and creep forces, refer to Andersson and
Knothe.2.6, 2.7

2-12
2.4.2.5 Gravitational Forces
The wheel/rail contact also has to transfer the weight of the vehicle into
the track structure. The sum of the vertical forces Q is equal to the axle load
(figure 2-14). Because of the profiled wheels, the contact patches are inclined
at a general contact angle γ. This results in the generation of a lateral compo-
nent to the vertical load at each contact patch. When the wheelset is placed cen-
trally on the track, the lateral components at each contact patch are equal and
opposite. However, this situation changes when the wheelset is displaced from
the track center (figure 2-14). The angle of the contact patches on each wheel can
be different and result in a net force acting on the wheelset.

Figure 2-14: Lateral Components of Normal Forces on the Wheelset

Generally, gravitational forces can assist in vehicle guidance although this
effect is not of great significance and is not identified as being the cause of
significant track or vehicle damage.

2.4.3 Flange Forces

The flange makes contact, laterally, with the rail under the following
conditions:
• When the wheelset runs with a significant angle of attack for a significant
distance. Lateral creep forces are generated as discussed previously and
as shown in figure 2-13. These forces tend to force one flange into contact
with the rail.
• When a vehicle runs through a curve at speed unbalance (see
section 2.8.1). In this instance, the wheelset can be forced into contact
either with the high or low rail.
• When a vehicle/wheelset momentarily loses its steering ability or
negotiates a track discontinuity where it is unable to steer (typically
turnouts)

2-13
 Figures 2-15 and 2-16 show typical forces occurring when flange contact
occurs.

Figure 2-15: Creep Force and Flange Force

Figure 2-16: Forces on a Wheelset with Flange Contact

Generally, the combined action of flange forces, acting in concert with the
lateral creep force are responsible for the typical wheel and rail damage seen in
sharper curves of heavy haul lines (figure 2-17). This occurs when the wheelset
flanges under an angle of attack as discussed in the first bullet of section 2.4.3.

Figure 2-17: Forces on a Wheelset in Flange Contact and Resultant Wheel and Rail
Damage

2-14
2.5 Introduction to Damage Criteria
Wheel/rail forces between the vehicle and track are transmitted across the
contact patches. They create stresses and creepages across these contact
patches and can generate one or a combination of the following at and/or
adjacent to the contact patch:
• Wear
• Material flow
• Surface and subsurface flow
The science of prediction of these parameters in the railway environment is
still in a developmental stage. The status of this science is provided in chapter 3.
Vertical loads and stresses are driven by axle load and vertical dynamics. These
are managed through load control, minimization of vertical dynamics, and wheel
and rail profile design and control to maximize the area of the contact patch and
minimize contact stresses. Wheel profile design is addressed in section 2.7.
Lateral loads are produced by the steering properties of the vehicle and are
a current strong driver of heavy haul bogie design, particularly on moderate to
severe curvatures and in severe, cold-climate conditions. Consequently, much
emphasis in this chapter is placed on bogie design and steering properties
(reference section 2.6).

2.6 The Railway Vehicle and the Dynamics of Assemblies of Two

Wheelsets
Any practical railway vehicle requires at least two wheelsets (four wheels) to
support and guide the vehicle. The manner in which these wheelsets are coupled
to the vehicle has a significant influence on vehicle cost, the performance of
the rail and the wheel, and the guidance and dynamics of the vehicle. Although
there is a strong relationship between vertical and lateral dynamics, vehicle
suspensions are generally discussed separately in terms of their vertical and
horizontal properties:
• The horizontal suspension. This provides lateral and longitudinal coupling
between wheelsets and between these wheelsets and the wagon body.
This suspension determines the tracking and curving performance of the
vehicle.
• The vertical suspension. This carries the load, attenuates dynamic forces,
and ensures adequate vertical wheel/rail loading at each contact patch to
ensure effective vehicle guidance while avoiding derailment.
Vehicle dynamics cannot be discussed without considering the properties of
the track. Hence, track geometry and track stiffness are also included in this
generic discussion of railway suspension systems.

2-15
2.6.1 Vertical Suspension Considerations
The purpose of the vertical suspension is threefold:
A. Attenuation of Vertical Vibrations
When moving forward on the track, the vehicle experiences vibrations of
varying frequencies. These vibrations excite the various modes of the vehicle
structure, body, and the payload. The dynamic body modes are: Bounce, roll,
pitch, nosing, and sway, as well as coupled modes of these basic motions. Some
of the mechanisms exciting the wagon body are:
• Long wavelength, track alignment irregularities in the vertical profile and
track twist. These irregularities typically result in vehicle input
frequencies between 0.5 and 30 Hz.
• Long wavelength, track stiffness variations are also present and activate
the vehicle in similar modes and frequencies as the alignment
irregularities.
• Short wavelength, consistent stiffness variations associated with local
sleeper support, result in vehicle input frequencies up to 40 Hz.
• High-frequency impacts at rail discontinuities (P1 forces) often excite the
vehicle body vertical modes to induce the so-called P2 lower frequency,
reaction forces.
B. Equalization of Wheel Loads by the Vehicle Suspension
As the vehicle is supported on a minimum of four contact patches (two wheel-
sets) on perturbed or twisted track, it is generically a statically indeterminate
structure similar to a four-legged table on an uneven floor. As section 2.2.1
states, sufficient vertical load is required across the contact interface for effective
guidance. The vertical suspension stiffness must thus prevent unacceptable
wheel unloading on twisted track.
C. Attenuation of Vertical Vibrations to the Track Structure
As a result of vertical, vehicle dynamics, dynamic loading is transmitted from
the vehicle through the wheel into the track substructure. Track elements, such
as the rail, the rail pads, the sleepers, as well as the ballast and the subballast
layer, are thus directly influenced by the dynamic performance of the railway
vehicle.
Typical exciting mechanisms are:
• Vehicle body dynamics in the frequency range between 1 and 30 Hz
• Out-of-round wheels (10 to 20 Hz)
• Wheel flats (10 to 20 Hz)
• Rail irregularities, such as rail joints and rail burns
Constraints that challenge the vehicle dynamist to design the optimum vehicle
suspension are typically:
• Limit on the minimum vertical, vehicle stiffness because of a limit on the
coupler height differential between adjacent vehicles in the tare and
loaded condition

2-16
 • Volume occupied by, as well as the stresses within, the suspension
• Coupled dynamic modes, particularly those of:
n The coupled excitation of wheelset and the wagon body (one inducing
resonance of the other)
n The excitation of the wheelset and the wagon body by regular track
features, particularly rail joints and deformation of the track structure
åßinduced by the action of the vehicle at rail joints
• Initial cost of the suspension
• Maintenance cost of the suspension
Similarly, the track engineer is limited in what can be done to optimize the
track structure. Typical constraints are:
• The cost of track components (example: ties, rail pads, under-tie pads)
• Cost constraints on the amount of ballast and formation material and
other geotechnical materials that could be used (example: ballast mats,
micro-piles, rammed aggregate piers, etc.)
• Track construction and track maintenance costs
Track dynamics analysts often include the so-called Hertzian stiffness in
their analysis. This is the vertical stiffness attributed to the deformation of the
wheel and rail under load. It is a high-order stiffness and is associated with
high-frequency vibrations and impacts. These are mainly of concern to track
engineers, and hence this effect is included under the section on track support
structures.
In its simplest form, the suspension of a railway vehicle comprises four spring
assemblies vertically coupling the four journal bearings on two wheelsets to the
body (see figure 2-18). The four springs can be designed to fit within the space
for relatively light vehicle loads. However, as the vehicle becomes heavier and
larger, the following factors come into play:
• The ability to accommodate track twist by means of spring deflection
clashes with the demands on coupler height differential limits between a
loaded and an empty vehicle.
• The available space for springs and dampers in the region of the journal
bearing is limited.
• As the carrying capacity of the vehicle increases, the wheel base
increases. This leads to increased demands on vertical deflection to
accommodate track twist.

2-17
 Figure 2-18: Simple Vehicle Suspension Arrangement

2.6.2 The Freight Vehicle Bogie

A solution to some of the problems discussed above is the railway bogie or
truck. The bogie is a combination of a minimum of two wheelsets within a
suspension structure that is pivoted beneath the vehicle body as shown in
figure 2-19. A minimum of two bogies is fitted to a vehicle. The bogie is the
equivalent of a short-wheelbase vehicle with a limited, but adequate, vertical
spring deflection to accommodate track twist. In addition, the carrying center
plate is of limited diameter. This coupling can be designed to permit additional
track twist by providing sufficient side bearing clearance.

Figure 2-19: Basic Railway Bogie

The bogie has become standard equipment under railway vehicles. Two basic
bogie designs are discussed below.

2-18
2.6.2.1 The Rigid Frame Bogie
The rigid frame bogie acts, vertically, very much like the model of the simple
railway vehicle described above. As indicated by the name, the single, rigid bogie
frame is typically in the form of a rigid “H” shape as figure 2-20 shows. The load
of the vehicle body is transferred from the center pivot through the “H-frame” to
the springs placed above the journal bearings. These springs form the vertical
suspension and cater to all the requirements for the suspension as listed above.
This type of bogie has not found favor in heavy haul operations for the follow-
ing reasons:
• Space constraints for springs with adequate carrying capacity and
deflection in the region of the journal
• Cost of providing four separate spring/damper systems on the bogie
• Cost of the H-frame from a manufacturing complexity and tolerance point
of view
The three-piece bogie has proven superior in all these respects.

Figure 2-20: Rigid Frame Bogie

2.6.2.2 The Three-Piece Bogie

The three-piece bogie, as implied by the name, is comprised of two side frames,
each resting in a longitudinal orientation, on the journals of the wheelsets.
Figure 2-21 shows a typical three-piece bogie. The side frames support a cross
member (the third piece) termed a bolster. The bolster is fitted with a center
pivot that couples the bogie to the vehicle body. The three pieces, two side
frames and a bolster, are each simply supported beams. This makes the bogie a
more statically determinate structure than the rigid frame bogie as the structure
can articulate under conditions of track twist with limited loss in vertical wheel
load. The advantages of this structure for vertical suspension are:
• It efficiently accommodates track twist.
• Suspension springs are limited to two nests, offering cost advantages with
respect to the number of suspension elements.

2-19
 • Suspension springs are in a region of the structure where more space is
available than at the journal boxes.

Figure 2-21: A Typical Three-Piece Bogie

A disadvantage of the three-piece bogie is that it may provide less than
adequate wheelset guidance as the side frames and wheelsets can lozenge
or warp (figure 2-22). Friction wedges are used in the suspension to dampen
both vertical and lateral vibrations, as well as to provide a degree of “squaring”
against this lozenging (warp) action. Vehicle dynamics and wear can, however,
compromise this function.

(a) Bogie warp in curves (b) Bogie warp when hunting

Figure 2-22: Warp of Three-Piece Bogies

2-20
 Consequently, three-piece bogies are often characterized by:
• Poor curving performance. As figure 2-22(a) shows, the bogie not only
constrains the wheelsets parallel to one another (as opposed to
permitting radial alignment of the axles) but also, through the warp
deflection of the side frames, sometimes allows the wheelsets to take up
exaggerated angles of attack. This induces high curving forces, flange
wear, and rolling contact fatigue on the wheel treads.
• Poor hunting stability. Figure 2-22(b) shows poor hunting stability as the
wheelsets may not be held “square” to one another in a sufficiently stiff
manner.
Despite these drawbacks, three-piece bogies continue to be the most com-
mon type of bogie design for heavy haul freight wagons. However, over the past
two or three decades, researchers and vehicle designers have developed a far
better understanding of vehicle dynamics. They now have the ability to develop
accurate computational models that has led to the development of a number of
advanced bogie designs.

2.6.3 Dynamics of Railway Vehicles on Tracks

The dynamic behavior of a railway vehicle influences vehicle/track forces,
derailment stability, and ride quality. Dynamic analysis can provide a guide to
optimizing vehicle and track design parameters and maintenance limits.
Traditionally, railway vehicle dynamic analysis has focused on the railway
vehicle while assuming a rigid track structure. Today, with increased
computational capacity, the analysis can include the stiffness and damping
properties of the track. The vehicle and track can be analyzed as a system.
Vibrations of the vehicle are transmitted to the track and excite vibrations of the
elastic track structure, which can couple with vibration modes of the vehicle,
both vertical and lateral.

2.6.3.1 Analyzing the Vehicle/Track System

The focus of the analysis of railway vehicle dynamics can be to improve
ride quality, to prevent vehicle hunting, to reduce deformation of the track, to
minimize damage and wear to both vehicle and track components, and to en-
sure running safety. Therefore, it is desirable to systematically investigate the
dynamics of a vehicle and the track from the entire vehicle/track system
viewpoint.2.6, 2.35, 2.36, 2.37
Figures 2-23 and 2-24 illustrate the components of a three-dimensional,
vehicle/track coupled model for a typical freight car with three-piece bogies on
ballasted track, in which the vehicle subsystem and the track subsystem are
spatially coupled with wheel/rail interaction.

2-21
Figure 2-23: Model of Vehicle/Track Coupled Dynamics for Freight Car
with Three-Piece Bogies (end view)

Figure 2-24: Model of Vehicle/Track Coupled Dynamics for

Freight Car with Three-Piece Bogies (elevation)

2-22
 The axle box suspension is modeled typically as shown in figure 2-25.

(a) (b) (c)

Figure 2-25: Modeling of the Axle Box Connection between Side Frame
and Wheelset
The track structure is often modeled as two continuous Bernoulli-Euler beams
supported at the sleeper by springs and dampers representing the rail pad. The
ballast and the subgrade are modeled. Three Degrees Of Freedom (DOF) are
considered for both left and the right rails. A five-parameter model of ballast
under each rail supporting point is used (figure 2-26), based upon the hypothesis
that the load transmission from a sleeper to the ballast follows a conical path.
Only the vertical motion of the ballast mass is taken into account.

Figure 2-26: Modified Model of the Ballast

The wheels and the rails are coupled using a wheel/rail interactive model as
shown in figure 2-27. This model considers three kinds of rail motions in the
vertical, lateral, and torsional directions. This coupling model eliminates the as-
sumption that the wheel and the rail are in contact all the time and may provide
a new path to simulate the dynamic process of derailment.

2-23
 Figure 2-27: Wheel/Rail Coupling Model
Utilizing the vehicle/track coupled model, the dynamic behavior of the en-
tire vehicle/track system can be analyzed, including vehicle hunting, curving
performances, and vehicle/track interactions.

2.6.3.2 Vehicle Hunting - General

In section 2.3, the self-centering phenomenon of a conical wheelset has been
described. This characteristic can lead to vehicle hunting, which results in
sustained lateral motion of a wheelset or vehicle above a critical speed. To
investigate vehicle hunting, the equations of motion are derived including all
effects disregarded by Klingel (see section 2.3.6), specifically the vehicle and
component masses and suspension and the correct wheel/rail kinematics and
mechanics.
Typically, as shown in figure 2-28:
• Below a certain speed called the “critical speed” of the vehicle, the
hunting amplitude decreases.
• Running exactly at the critical speed, the amplitude of the motion remains
constant.
• If the critical speed is exceeded, the amplitude of the hunting motion
increases.

Figure 2-28: Lateral Displacements of the Wheelset for Speeds Relative to the
Critical Speed
Figure 2-28 is, however, only valid for a completely linear system. In reality,
several nonlinearities are in the railway system, both regarding the wheel/rail
contact kinematics and mechanics and the vehicle suspension. The real hunt-
ing behavior commonly occurring can rather be described by a so-called limit

2-24
cycle diagram for nonlinear systems (figure 2-29). For stability studies of vehicle
systems in practical engineering, the nonlinear critical speed vB is generally used
as an important index to evaluate the vehicle performance.

Figure 2-29: Generic Limit Cycle Diagrams for a Railway Vehicle

In a bogie vehicle, basically three types of hunting motion can arise:
• Wheelset hunting where one wheelset performs the hunting motion
• Bogie hunting where the bogie executes a hunting motion
• Car body hunting where the car body is excited in one or more of its
natural resonant modes. Car body hunting is often a type of resonance
phenomenon, where the Klingel hunting frequency given mainly by
vehicle speed and conicity coincides with one of many frequencies of the
car body.
Hunting is a function of the wheel/rail geometry, the suspension and the
masses, and the inertias of the vehicle. Since the mass and inertia (and in
most cases, the suspension stiffness and damping of the freight wagons) will
significantly change from empty to loaded wagon, the type of hunting motion
observed usually differs between an empty and a loaded wagon. Since the
stiffness values between axle box and bogie frame are lower in an unloaded
vehicle, the risk for wheelset or bogie hunting is higher.
Loaded car body hunting (termed “loaded car hunting” in North America) can
induce high wheel/rail forces, particularly longitudinal forces, as the wheelset is
forced into alternating flange contact under high vertical loads.
Also note that, traditionally, vehicle hunting performance has been analyzed
assuming a rigid track. When an elastic track is modeled, up to 5% lower critical
speeds are predicted. This is due to the effect of track vibrations, especially the
lateral and torsional vibrations, of the left and right rails. Although the difference
of vehicle critical speeds based on the two track models is not too large, spe-
cial attention is worth being paid to the phenomenon because the conventional
vehicle dynamics regarding the track as a rigid base overestimates the stability
of a vehicle, which could potentially increase the possibility of fatalness in the
design stage of the vehicle.

2-25
2.6.3.3 Hunting Characteristics of Empty and Loaded Wagons
To further understand the effect of wagon load on hunting characteristics,
note that the conventional three-piece bogie utilizes the suspension wedges to
provide a stabilizing coupling between two coupled wheelsets. For three-piece
bogies using load-sensitive friction damping, the bolster suspension contains an
inner coil spring and an outer coil spring with different free heights, respectively
(figure 2-30).
For the empty wagon, only the outer coil spring is compressed with a small
stiffness and large static deflection. For the case of the loaded wagon, as the load
increases, the outer coil spring is being compressed gradually to the position
of the free height of the inner coil spring first (first stage), and then both the
inner and outer coil springs bear the load together (second stage). This results
in a bilinear force characteristic of the suspension as shown in figure 2-31. The
symbols f1 and f2 are the static deflections for the empty and the loaded wagons,
respectively. fT denotes the static deflection at the transition from the first stage
(stiffness K1) to the second stage (stiffness K2). In general, the stiffness K1 rep-
resents the load capacity for the tare (empty) condition, while the stiffness K2 is
for the loaded wagon condition. Meanwhile, the larger vertical stiffness of the
coil spring in the loaded condition causes an increase of the lateral and warp
stiffness, enhancing the elastic constraint capacity of the side frame to the bol-
ster, thus improving the running stability of the three-piece bogie.

Figure 2-30: Friction Damping

Figure 2-31: Bilinear Force Characteristic of Friction Damping

2-26
 Friction damping in the spring nest contributes the main damping effect
(figure 2-32). According to the kinematics of the components, two different
working states exist for the damping system (figure 2-33). From empty to loaded
conditions in both states, the normal forces (F1, F2) in friction pairs, as well as the
support force (P), increase, leading to larger lateral and vertical friction forces.
The contact friction further strengthens the damping effect, which could restrain
the hunting oscillation.

Figure 2-32: Relative Motion between Side Frame, Wedge, and Bolster

Figure 2-33: Force Conditions of Wedge

Taking the same wagon mentioned above, for example, the calculated
nonlinear critical speeds for the tare and the loaded conditions are, respectively,
77 kph (48 mph) and 113 kph (70 mph). Clearly, the hunting stability of the
empty wagon is worse, and its critical speed has reduced by 31.9% with respect
to that of the loaded wagon.

2.6.3.4 Vehicle Curving Performance

More than one wheelset is required to be coupled in the horizontal plane to
achieve wheelset stability at any practical vehicle speed. The manner in which
this is done influences the ability of coupled wheelsets to negotiate curved track.
To illustrate the basic curving properties of a wagon bogie,2.11, 2.12 consider two
example bogies, one with a high bending stiffness and the other with a low
bending stiffness.
A high bending stiffness implies that both wheelsets remain essentially paral-
lel to one another, and hence may not attain a radial position in a curve. There
is thus a limit on the ability of the bogie to negotiate sharp curves without
flange contact. This limit is a function of track gage, bogie wheelbase, wheelset
conicity, gage clearance, and bogie rotational resistance. Curving without

2-27
flange contact is shown in figure 2-34. The “clockwise” moments resulting from
longitudinal creep are in balance with the “anti-clockwise” moments resulting
from lateral creep. Hence, the bogie is kept in equilibrium.

Figure 2-34: Curving without Flange Contact

The “anti-clockwise” moments on the bogie due to the lateral creepage
(resulting from the angle of attack of the wheels) are larger than those
“clockwise” moments that can be generated by the longitudinal creepage, which
must thus be supplemented by a flange force. Once the wheelset is steered
by the wheel flange, severe rail side gage face wear may happen, as shown in
figure 2-35.

Figure 2-35: Flange Contact in Curves

Bogies with a low bending stiffness, generally termed “steering bogies,” use
the longitudinal creep forces generated between the wheelset and the rail to
deflect the longitudinal springs, which creates the bending stiffness. This
permits the wheelsets to align to an almost radial position to the curve, as
figure 2-36 shows. The lateral creep forces are reduced to almost zero, eliminat-
ing the flange force and its effect on the rail.
2-28
 Figure 2-36: Radial Alignment in a Curve

2.6.4 Advanced Freight Wagon Bogie Designs

New advanced design bogies have been introduced to address the
disadvantages of the three-piece bogie with the goal of improving vehicle
stability and curving performance.

2.6.4.1 Premium Bogies

So-called premium bogies have been developed in response to the desire
to improve three-piece truck performance by preventing the frame “warping”
behavior. In one design, called a frame-braced bogie, a pair of diagonal braces
are connected between the side frames and clamped to each other at the cen-
ter, providing the desired shear (or anti-warp) stiffness. Elastomeric pads are
located between the bearing adapters and the side frames to provide controlled
resilience in the horizontal directions. This adaptation has been applied as a
retrofit on many bogies (trucks) and also as original equipment on many
thousands more. A schematic representation can be seen in figure 2-37.
Warp stiffness can also be obtained by using a spring plank or
transom (figure 2-38). The spring plank is designed to constrain the two side
frames against warping through the use of some form of fixture between the
plank and the side frame. This fixture can take the form of lugs, rivets, or lock
bolts.
Variations of this design can also allow the side frames to rock or execute a
swinging/pendulum motion. This can assist in laterally “uncoupling” the wagon
body mass from the wheelsets, improving loaded wagon hunting stability.

2-29
 Figure 2-37: The Frame-Braced Truck

Figure 2-38: Bogie Fitted with a Spring Plank (Transom)

An advantage of the increased warp stiffness provided by spring planks
or frame braces is that a degree of resilience in axle longitudinal and lateral
guidance may now be introduced at the bearing (figure 2-39).

2-30
 Figure 2-39: Adapter Pad Fitted to a Premium Bogie
This feature provides some freedom for the wheelsets to align:
• Into a parallel position relative to one another to compensate for
manufacturing tolerances in the bogie
• Into a partially radial position in shallow curves to improve curving

2.6.4.2 Self-Steering Bogies

Self-steering bogies use the longitudinal, creep forces generated between the
wheelset and the rail to deflect the longitudinal springs at the wheelset jour-
nal. These springs create the bending stiffness of the bogie. This permits the
wheelsets to align to an almost radial position in a curve as figure 2-40 shows.
The lateral creep forces are thus reduced to almost zero, eliminating flange
forces. Self-steering bogie designs take advantage of the steering ability of
wheelsets with profiled wheels.

Figure 2-40: Radial Alignment in a Curve

2-31
2.6.4.2.1 Interaxle Shear Stiffness
Research has shown that a certain amount of interaxle shear and bending
stiffness is required for adequate dynamic vehicle stability.2.9, 2.10, 2.11, 2.12
Conventional bogies with a rigid bogie frame use the shear stiffness of the bo-
gie frame to obtain interaxle shear stiffness. However, in the three-piece bogie
arrangement, the shear stiffness of the three-piece frame is inadequate for
optimal stability. Furthermore, if the frame is used for the transmission of shear
forces between the wheelsets, a high wheelset yaw constraint is required. This
is contrary to the requirement of steering bogies.
Consequently, to optimize the curving and stability performance of three-piece
bogies, interaxle shear stiffness, independent of the wheelset yaw constraint,
is required. In this case, the longitudinal and lateral stiffness of the primary
suspension can be selected to best suit curving performance, stability, and ride 3
quality. 3 N P₀,max =
P₀(ξ,η) = 2 A √ 1 - ( aξ )² - ( ηb )² ,
e
2

2.6.4.2.2 Linear Wheelset Yaw Constraintτ₁,max = 0.3 ∙ p₀ (32) = (1

at z = 0.78a.
A study of the quasi-static deflection of the wheelset of a two-axle bogie on
curved track in the off-flanging mode shows that the wheelset deviations from
the pure rolling position are independent of the interaxle shear stiffness ifr the
l
Rw δ
y = 0 to
λ
interwheelset bending stiffness (or wheelset
γ = yΔr yaw
γe =constraint)
R w - RR is selected
Rc γ
comply with the equation, which relates the bending stiffness to the lateral pro-
file stiffness of the wheels.2.11 This equation is:
f Effec
(2.8) kb² = 11 kp l²
f22 [ R-l ]
Δr = r0 R + l - 1

where
k = longitudinal axle box stiffness
2b = distance between axle boxes
f11 = longitudinal creep coefficient
f22 = lateral creep coefficient
kp = profile stiffness
2l = distance between wheel/rail contact points
If this relationship is adhered to, high, interaxle, shear stiffness can be
provided for optimal hunting stability without adversely R R δ 2 l λthe
Δ r γ yaffecting π off-flangew e
curving performance of the bogie.2.11

2.6.4.2.3 Variable Wheelset Yaw Constraint

The maximum operating speed of self-steering bogies with a linear wheel-
set yaw constraint, which complies with equation 2.8 is generally limited to
120 kph (74 mph). For higher operating speeds, the bending stiffness can be
supplemented by a so-called “Variable Yaw Constraint Suspension System.” This
system is a passive, elastic, suspension element that provides a bending stiffness
between the wheelsets that is varied by the creep forces generated in such a
manner that off-flange curving is still maintained in all main line curves.

2-32
2.6.4.3 Developments in Self-Steering, Three-Piece Bogies
The abovementioned concepts have led to the development of self-steering,
three-piece bogies such as the cross-anchor bogie, the bissel (or wishbone)
frame bogie, and the radial-arm bogie. These bogies are briefly described below.

2.6.4.3.1 Cross-Anchor Bogie

The cross-anchor, three-piece bogie design (figure 2-41) retains the three-piece
frame arrangement as well as the load-sensitive vertical and lateral damping, but
introduces rubber shear pads at the journal boxes for a controlled yaw motion of
the wheelset and provides additional interaxle shear stiffness independent of the
shear constraint provided by the three-piece bogie frame. This interaxle shear
stiffness is obtained by diagonally linking the journal boxes by means of cross
anchors.

Figure 2-41: Cross-Anchor Bogie

2.6.4.3.2 Bissel Frame Bogie

The bissel or articulated three-piece bogie (figure 2-42) incorporates a pair of
steering arms interconnected by an elastomeric connection located in the center
of the bogie into the conventional three-piece bogie. Hence, this bogie design
has a reduced interaxle bending stiffness to enhance the curving performance
and an increased interaxle shear stiffness to provide adequate vehicle stability.

Figure 2-42: Bissel Frame Bogie

2-33
2.6.4.3.3 Radial-Arm Bogie
The radial-arm bogie (figure 2-43) is a further development of the cross-anch
or bogie. Like the cross-anchor bogie, it connects the two wheelsets of the bo-
gie in shear. This is not achieved, however, by cross anchors, which have to be
fitted diagonally through the bolster and be connected to subframes, but by ra-
dial arms that are positioned at the outside of the side frames. For three-piece
bogies, the radial steering arms form an integral part with the bearing adapters
that rest on the package-type journal bearings.

Figure 2-43: Radial-Arm Bogie

Various aspects relating to the design and development of the radial-arm,
self-steering, three-piece bogie by Scheffel, as well as the development of an
optimum wheel profile for self-steering bogies under heavy axle load conditions,
have been described in literature.2.11, 2.12, 2.13, 2.14, 2.15, 2.16, 2.17 The radial-arm, self-steer-
ing bogie has also found its application in 215.5 kg (30 T) axle load heavy haul
operations on standard gage track.2.14

2.6.4.4 Body-Steered Bogies

Ever since railroads began in the early 19th century, there have been proposals
for bogie designs in which the wheelsets are steered into radial alignment with
the track.2.18
A concern with steering bogies is to reconcile the improved steering action
permitted by radial alignment of the axles in conjunction with wheelset conicity
with that of stable running (hunting stability).
Body-steered trucks use a separate mechanism that determines the steered
position of the wheelsets.2.19, 2.20, 2.21
This mechanism can come in different forms:
• Some mechanisms utilize the roll angle of the vehicle body in a curve to
cause the wheelsets to be displaced into a nonparallel configuration.
These mechanisms suffer from the difficulty that they depend upon a
specific unbalance in the curve to create the wheelset alignment.

2-34
 • Similarly, there are mechanisms that create the desired wheelset
alignment by reacting to the lateral displacement of the vehicle body
relative to the truck frame when traversing a curve. Again, this style of
mechanism depends upon the unbalance in a curve to create the steered
alignment, and at any other than balance speed in a curve, the steering
will be incorrect.
Both mechanisms mentioned above also suffer from the fact that any roll
or lateral disturbance experienced in tangent track will create a steering
response, which could be detrimental to the vehicle’s behavior.
• Another class of mechanism that has been proposed utilizes sensors and
powered actuators to achieve the desired steering effect. The curvature
of the track is detected by a sensor on the truck, this sends a signal to an
actuator, and the actuator moves the wheelsets into the desired position.
• A further type of body-steered truck has been developed in which the
degree of curvature of the track is determined by the yaw angle between
the bogie frame and the vehicle body. The yaw angle of a bogie in a wagon
is directly proportional to the degree of curvature of the track. By
connecting a mechanism between the truck frame and the vehicle body,
which moves in response to this swivel angle and acts upon the wheel
sets, it is possible to effect a displacement of the wheelsets relative to the
bogie frame such that the wheelsets must always be in perfect radial
alignment with the curvature of the track. Such a mechanism is not
dependent upon the speed of the vehicle – the yaw angle is independent of
that factor – so the steering will be the same for all conditions of over-
balance and under-balance speed in a curve. This type of mechanism also
requires no external power source or electronic device in order to be
effective. Even given this definition of steering, it is possible to have two
distinct types of bogies – ones in which the wheelsets are constrained
from moving relative to the bogie frame, and thus the mechanism must
overcome this constraint – and ones in which there is no constraint against
the angular motion between the wheelsets and the truck frame, so the
mechanism simply guides the wheelsets into the correct position and
constrains them to stay there. A schematic diagram of a body-steered
bogie is shown in figure 2-44.

2-35
 Figure 2-44: A Steered-Axle Railway Truck
This mechanism allows both axles to be steered to equal and opposite angles
relative to the transverse center line between them and that there must be little
or no lateral (shear) motion between them. With this condition met, the wheels
will follow one another on the same path through a curve.
There are a number of design considerations for a body-steered bogie:
• The axles must be able to move freely relative to the bogie frame for
radial alignment under the vertical load at the bearings.
• The introduction of steering errors by the mechanism chosen must be
avoided. The yaw angle of the bogie under the wagon body is used as the
input to determine the amount of yaw of the axles. There are
nonlinearities in the system that need careful design.
• A coupling between wheelset yaw and the wagon body yaw could
introduce instability.
When negotiating a transition curve or spiral into a curve, the input to the
body-steered system will not reflect the true local radius of curvature at each
of the bogies, so there will be errors in the steering. The leading truck will be
under-steered and the trailing truck over-steered until the vehicle is fully onto
the simple curve.
These considerations can be overcome. Tests indicate steered-axle bogies can
reduce rolling resistance on tangent as well as curved track, improve hunting
stability, and improve derailment stability.

2.7 Wheel/Rail Design Strategies

Wheel and rail profiles have significant effects on vehicle performance and
wheel/rail lives. Contact conditions are closely related to the profile shapes and
their compatibility. During operation, contact occurs on specific portions of the
wheel and rail, not over the complete surface. It must thus be concluded that
wear must change the shape of both wheel and rail. This section describes these

2-36
changes in shape, suggests that any change in shape introduces deleterious
conditions, and offers means to best minimize and manage this change in shape.

2.7.1 Basic Wheel and Rail Contact Patterns

Wheel/rail contact generally occurs under three main conditions:2.22
• The wheel tread contacts the rail on the center region of the head. This
type of contact generally occurs on tangent track and on curves with very
large radii (figure 2-45a).
• One wheel contacts the high rail on the rail “shoulder” (close to, but not
on, the gage corner). The other wheel contacts the low rail on the center
region of the head. This type of contact generally occurs in shallow
curves (figure 2-45b).
• One wheel contacts the high rail on the gage corner and low rail contact
is made on the center of the rail head. This type of contact generally
occurs on curves with small radii (figure 2-45c).

(a)

(b)

(c)
Figure 2-45: Typical Wheel/Rail Contacts under Normal Conditions
Variations in (a) wheel/rail profile designs, especially in grinding tem-
plates, (b) actual ground rail shapes, (c) wheel/rail wear, (d) track geometry
perturbations, and (e) vehicle performance can cause more (and more complex)
actual wheel/rail contact conditions in service.

2-37
 Wheel/rail contact may also be described in the manner shown in figure 2-46
and related to three main regions on the wheel tread and rail head.2.22

Figure 2-46: Three Contact Regions

Under normal operating conditions, wheel and rail profile contact combina-
tions can generally be categorized as table 2-1 shows. The distributions of these
combinations depend on the new wheel and rail replacement rates, rail grinding
intervals, and wheel and rail wear conditions. The optimized wheel/rail contact
system should make variations in the contact conditions in as narrow a band as
possible. This can be achieved through carefully designed wheel and rail profiles
and maintenance strategies.
Combination Types Wheels Rails
1 New New
2 New Differing levels of worn rails
3 Differing levels of worn New
wheels
4 Differing levels of worn Differing levels of worn rails
wheels
5 New Newly ground
6 Differing levels of worn Newly ground
3 N wheels P₀,max = 32 - NA
P₀(ξ,η) = 2 Ae √ 1 - ( aξ )² - ( ηb )² , e
Table 2-1: Complicity of Wheel and Rail Combinations
()=(
3
2 1
tan(δ)
+ μ *
-μ
tan(δ) )
τ₁,max = 0.3 ∙ p₀ at z = 0.78a.
2.7.2 Influence of Wheel/Rail Profile Compatibility on Vehicle
Curving
lr0 v
For a wheelset with a rigid axle
Rwto δ properly negotiate
r l a curve, theλ wheel
= 2π contact- f = 0
Δr γe = y = 0 γ λ
ing the outer
γ = y rail must roll onRaw -larger
RR rolling radius
Rc γ than the wheel contacting
the inner rail (described earlier in section 2.3). Equation 1 relates the difference
in rolling radius Δr on a wheelset in relation to the geometry of the wheelset:
f Effective Conicity (λ) = RRD
kb² = 11 kp l²
(2.9)
f22 [
Δr = r0 R + l - 1
R-l ] 2y

Where R is the curve radius measured from track center line, l is half distance
of two rail contact points (see figure 2-7), and r0 is the nominal wheel radius.
γ = rolling radius differentia
Conditions for the onset of on-flange curving depend on the curving perfor-2y
mance of bogies, track gage, and wheel/rail clearance.

2-38
2.7.2.1 Wheel/Rail Contact Conditions in Tight Curves

2.7.2.1.1 High Rail Contact

In tight curves, the wheels on the high rail are generally in flange contact to
achieve the required Δr. Depending on the shape of the wheel/rail profiles, four
types of contact can take place (figure 2-47):
(a) Strong, two-point contact with large rolling radius difference between
these two points and a gap between wheel flange throat and rail gage
corner [greater than approximately 0.4 mm (0.016 in.)]
(b) Strong one-point contact with small contact area
(c) Conformal contact with large contact area
(d) Near-conformal contact with small rolling radius difference between
these two points and a gap between wheel flange throat and rail gage
corner less than approximately 0.4 mm (0.016 in.)

(a) Strong, two-point contact (b) Strong, one-point contact (c) Conformal contact (d) Near-conformal contact
Figure 2-47: Contact Conditions on High Rail
Of these contact conditions, the strong, two-point contact and the strong
one-point contact with small contact area are not desired.
Strong, two-point contact has one contact point on the wheel tread and another
on the flange. A rolling radius difference is developed between these two points
that can reach 12 mm (0.47 in.) or more. This results in:
• Longitudinal creep forces acting in opposite directions, as figure 2-48a
shows. As a result, the wheelset steering moment is reduced, resulting in
higher angle-of-attack and lateral forces.2.23, 2.24
• Relative slip between these contact points and resulting high rate of wear
and material flow on the flange tip (figure 2-48b).

(a) Opposing Longitudinal Forces Cause Reduced Steering Moment (b) Plastic Flow at Flange Tip Contact
Figure 2-48: Detrimental Effect of Strong Two-Point Contact

2-39
 The causes of strong, two-point contact are generally:
• Incompatible wheel and rail profile designs
• New wheels contacting the worn high rail
• Wheels with high flanges contacting excessively worn rail (figure 2-49)
• Overgrinding the rail gage corner (figure 2-50a)2.1
• Recanting rails that have rolled on their rail seats in service without sub
sequent rail grinding (figure 2-50b)2.25

Figure 2-49: Excessive Rail Gage Face Wear Causing Worn

Wheel to Bear on the Gage Lip

High Rail

Figure 2-50: (a) Over-Cut Gage Corner (b) Recanting Rail without Rail Grinding
Single-point contact generally results in a small contact area at the high rail
gage corner and lower lateral forces because of a high rolling radius difference
between the wheels on the high and low rails and the resulting gravitational
force. However, high longitudinal creepages developed as a result of the high
radius differential Δr, in conjunction with high-contact stress areas, can cause
either, or both, a high rate of wear or surface cracks due to rolling contact
fatigue. Figure 2-51 shows an example of how the wear of a new high rail under
high-friction conditions (no gage face lubrication) resulted in the wear pattern
shown. The new rail profile was subjected to a strong one-point contact with
resulting high stresses and gage face cracking. High longitudinal tractions
caused a high rate of gage face wear until the wheel and the rail stabilized in
conformal contact conditions. Initial gage face lubrication, with an improved
conformal “match” of wheel and rail profiles with lower contact stresses allowing
the lubricating film to withstand the contact stresses, would have appreciably
reduced wheel and rail wear.
2-40
 Figure 2-51: Wear Pattern Producing Strong,
One-Point Contact
Near-conformal and conformal contact on the high rail are more desirable
contact situations to produce lower lateral forces, rolling resistance, and contact
stress. They can be achieved by:
• Designing compatible wheel and rail profiles
• Regular rail grinding and wheel/rail natural wear
Under heavy haul operation conditions, wheel and rail profiles that are
designed for near-conformal contact soon wear into full-conformal contact,
resulting in extended wheel and rail reprofiling intervals.

2.7.2.1.2 Low Rail Contact

The tread of the wheel should ideally contact the low rail in the center crown
region (figure 2-52).

Figure 2-52: Ideal Low Rail Contact Condition

This results in an optimal Δr on the wheelset and avoids “edge contact” on the
field and gage corners of the low rail. Edge contact induces material flow.
Low rail contact generally deteriorates with:
• Profile wear and material flow
• Rail rollover
• Increased gage (due to high rail side wear and flange wear and rail
rollover)
Low rail contact can move from the ideal, at the center of the rail head, to the
gage corner or to the field side of the rail.

2-41
 Contact at low rail gage corner (figure 2-53) can reduce the rolling radius
difference Δr on the wheelset, resulting in flange contact and higher lateral
forces. It can cause material flow on the low rail gage corner, reducing the gage
clearance and moving the contact position further to the low rail gage corner.
This contact condition can occur:
• When excessive material is removed from the field side of the low rail
• Under conditions of low rail roll

Figure 2-53: Contact toward the Gage Corner on the Low Rail
Contact on the field side of the low rail (figure 2-54) can:
• Reduce the rolling radius differential Δr on the wheelset, depending on
the (worn) shapes of the wheels
• Result in high-contact stress due to small contact area and cause metal
flow on the field side of both wheel and rail
• Cause the rail to roll if not adequately constrained (example: if secured
with cut spikes) due to the combination of vertical load and lateral creep
forces since contact is made far to the field side

Figure 2-54: Contact on the Field Side of the Low Rail

This contact condition generally occurs when:
• The low rail head has a worn flat top and makes contact with a hollow
tread (figure 2-55a).
• A flat low rail is recanted without grinding.
• There is a combination of wide gage and thin flange conditions.

2-42
 This problem is exacerbated when the field side of the wheel tread contacts the
center of the rail head, causing concavity of the head (figure 2-55).

Figure 2-55: Hollow Wheels Contacting the Concave Rail Head on the Low Rail

2.7.2.2 Wheel/Rail Contact in Shallow Curves

In shallow curves, flange contact generally does not occur or takes place
intermittently. The required rolling radius differential Δr is mainly dependent on
the slope of wheel tread and the rolling radius in the region near wheel flange
throat.
The rail shape can affect Δr. Figure 2-56 shows (in an exaggerated way) low
rail contact conditions. Contact at B would produce larger Δr than contact at A by
taking advantage of the wheel taper, assuming contact occurs toward the flange
on the wheel contacting the high rail.

Figure 2-56: Effect of Contact Position on Rolling

Radius Differential
3 N
3 N η ² P₀,max = -
P₀(ξ,η) = √ 1 - (The
2.7.3
2 Ae a ) - (Influence
ξ ²
b) ,
2 Ae
of Wheel/Rail Profiles on Vehicle Hunting
(2) =on 1both
Vehicle hunting can occur
τ₁,max = 0.3 ∙ p₀ at z = 0.78a.
3
( tan(δ) - μ
)
tangent track and shallow curves for both
+ μ * tan(δ)
empty and loaded cars. Lateral motion is limited only by the contact of the wheel
flange with the rail. The lateral force induced by hunting may cause wheel climb,
gage widening, rail rollover, track panel shift, or combinations
v0 of these effects.
lr0
r0 l
Rw δThe effective conicity λ
of wheel/rail contact
= 2π f
has considerable influence on the
=
γe = y = γ
R wvehicle
- RR Rc γ As wheelset conicity increases, λthe onset critical speed of
hunting speed.
hunting decreases. The simplified effective conicity is defined by equation 2.10:
Effective Conicity (λ) = RRD
(2.10)
Δr = r0
[
R+l -1
R-l ] 2y
where y is wheelset lateral shift relative to rail.

rolling radius differential

γ =2-43
2y
 Figure 2-57 shows two examples of low- and high-contact conicities together
with their related wheel/rail contact conditions. In the graphs of rolling radius
difference vs. lateral shift of the wheelset, conicity is defined as the slope of the
graph. The segment of this graph most influencing stability is the central portion
where the wheel tread is in contact with the rail head. Flange contact, reflect-
ed by the very steep portions of the graphs, has a lesser influence on hunting.
Consequently, the central section of the graph, before flange contact is made, is
used to compute the so-called effective conicity and to compute hunting stability.
In the example in figure 2-57, conformal contact (figure 2-57a) results in a
conicity of approximately 0.05 (1:20), whereas single-point contact conditions
shown in figure 2-57b result in a conicity of 0.65 (~ 1:1.5). The critical speed
for the wagon fitted with the low conicity profile (figure 2-57a) was in excess of
120 kph (74 mph), whereas the critical speed for the same wagon fitted with the
high conicity profile (figure 2-57b) was 75 kph (46 mph).

(a) Low Conicity (b) High Conicity

Figure 2-57: Effect of Wheel/Rail Contact on Conicity
The critical hunting speed is highly dependent on the vehicle/track character-
istics and wheel/rail profiles. Although the critical conicity value can be varied
by vehicle types, it has been recommended that this value should be lower than
0.35 in. heavy haul operation.
High conicity can also be caused by a rail shape with a flattened crown or
rail with metal flow at the gage corner and hollow worn wheels (figure 2-58).
The conicity increases with tight track gage. The tight track gage reduces the
clearance between wheel flange and rail gage, resulting in wheels contacting
the rail at the gage shoulder and corner. Small lateral shifts of the wheelset due
to track perturbations and vehicle dynamics can induce big changes in rolling
radius difference between the two wheels.

2-44
 Figure 2-58: Worn Wheel Profile Contacting a Rail with a Higher Gage Shoulder

2.7.4 Hollow and Asymmetrically Worn Wheels

2.7.4.1 Hollow Worn Wheels

Hollow worn wheels (figure 2-59) can cause a “false flange” to develop on the
field side of the wheel tread. When contact is made with the rail or other track
components (e.g., frogs), the convex shape of this “false flange” results in high-
contact stresses if the contact is on the extreme outside edge of the rail or frog.
Calculations show that the maximum contact stress can increase from less than
1500 N/mm2 for normal tread contact to 6000 N/mm2 resulting from false flange
contact. This can cause rolling contact fatigue on the surface of rails.2.27

Figure 2-59: Hollow Worn Wheel

Figure 2-60 shows the contact situations for a new wheel and a hollow wheel
on tangent track. At zero lateral shift between wheel and rail, both wheels
contact the rail on the center of the head. With a small lateral shift [less than 2 mm
(0.08 in.)], the new wheel tread still contacts the rail in the crown area, but the
hollow worn wheel throat will contact the rail toward the gage corner.
Consequently, small track irregularities can cause the wheel/rail contact point
to “move” laterally, very quickly. This effectively results in a high conicity and
can lead to hunting.

Figure 2-60: New and Hollow Worn Wheel Contact Conditions

2-45
 The transfer of contact between the wheel and elements of a switch can be
inhibited by the false flange that can be “jammed” between switch point and
stock rail. Instead of riding up to the stock rail, the false flange pushes the stock
rail laterally, as shown in figure 2-61, increasing the risk of stock rail rollover.2.28

Figure 2-61: Hollow Worn Wheel Increases the Risk of

Stock Rail Rollover

2.7.4.2 Asymmetrically Worn Wheels

An assymetrically worn wheelset has one wheel flange that is more worn than
the other wheel on the same axle (shown in figure 2-62). Assymmetrically worn
wheels are commonly associated with one wheel on a wheelset having a hollow
worn shape. Severe asymmetrically worn wheels can affect vehicle performance
on both tangent and curved tracks.2.29

Figure 2-62: Example of Wheel Profiles on an Asymmetrically Worn Wheelset

On tangent track, to reach a rolling radius equilibrium position, the
asymmetrically worn wheel with more flange and tread wear will tend to be in
flange contact with the rail, resulting in increased gage face wear on tangent
track. This condition may also increase the contact conicity by creating a large
rolling radius variation with small lateral shifts.
On curves, trucks with asymmetrically worn wheels can curve poorly in one
sense of curves, but well in the reverse sense because of the difference in rolling
radius difference generated between, particularly, the lead wheels when curving
in each sense. This is illustrated in figure 2-63. If RR1 and RR3 are the rolling
radii when contacting the high and low rails respectively then:
• When RR3 (low rail rolling radius) is greater than RR1 (the high rail
rolling radius), the wheelsets want to curve in a sense opposite to
the curve.
• When RR3 (low rail rolling radius) is smaller than RR1 (the high rail
rolling radius), the wheelsets want to curve in the “correct” sense in the
curve – and are “assisted” by the assymmetry.

2-46
 Figure 2-63: Bogie Curving Performance with
Asymmetric Wheels

2.7.5 Wheel and Rail Profile Design Strategies

Wheel and rail profile shapes need to be matched properly to ensure
advantageous contact conditions. Systematical analysis using contact mechanics
and vehicle dynamics needs to be conducted before adopting new wheel or rail
profiles. In service, worn wheels are generally removed and reprofiled back to
the originally designed shape, while rails are generally ground into shapes that
are different from the new original shape. The reason that rail is often ground to
a different shape in service is because this shape can be optimized with respect
to the specific position of the rail in the track (tangent vs. curve, high vs. low
rail). Rail grinding can assist vehicle curving requiring the generation of higher
rolling radius differentials Δr and lateral stability on tangent track requiring, in
turn, a lower Δr.
Wheel and rail profiles are generally designed in the form of a series of
circular arcs, cubic splines, or simply a set of x-y coordinates. Regardless of the
format, tangency of the adjacent segments and smooth connection of points are
required. No kinks in the contact zone should be allowed. The designed profiles
are not only being used in manufacturing new wheels and rails and rail grinding,
but are also being used in vehicle modeling simulations. Any discontinuities on
designed profiles may lead to distorted simulation results.

2.7.5.1 Wheel Profile Design

The design of a wheel profile2.22 must consider the operating environment
including the track conditions and the characteristics of vehicles. Wheel
profile designs for railways with mainly tangent track might be different from the
designs for railways with many curves of small radii. Profiles could also be
different for wheels on regular three-piece bogies that have relatively low warp
stiffness compared to those for steering bogies.

2-47
 Considering the effects of wheel/rail conditions on vehicle performance and
wheel/rail interface previously discussed, the following parameters should be
carefully considered in the design of wheel profiles:
• Sufficient flange face angle to reduce the risk of wheel climb on curves
and switches2.30
• Proper tread shape to compromise between curving and lateral stability
on tangent track
• Strong, two-point contact on high rail of curves to be avoided
• Strong, single-point contact with small contact area either at the wheel
flange root/rail gage corner or at the wheel tread/rail crown to be
avoided
• High-contact stress to be avoided
• Smooth transition between contact zones
For a completely new rail system, wheel and rail profiles should be designed in
pairs to achieve optimal contact conditions.
Consider the population of existing wheel and rail profile shapes when
designing a new wheel profile for an existing rail system to improve wheel/rail
contact conditions and wheel/rail lives. A smooth transition from existing con-
tact conditions to a new level of equilibrium needs to be carefully designed. The
wheel and rail profiling equipment capacity (such as wheel truing machines and
grinders) in the system should also be considered for designing the transition.
As an example, the optimization of wheel profiles in North American freight
service has been pursued in several stages in the past decades. In 1988, a new
wheel profile called AAR-1B was developed to produce less strong, two-point
contact on both new and worn rails.2.31 It replaced the AAR-1:20 wheel profile as
the North American freight service wheel standard. As the concept of conformal
contact is being adopted by freight service, efforts to further modify the AAR-
1B wheel have been made. Transportation Technology Center, Inc. (TTCI) has
developed the SRI-1A wheel profile and it is being tested in service.2.32 The
National Research Council of Canada (NRC) also developed the NRC-ASW
wheel profile. Both profiles are aimed to produce more conformal contact with
the worn high rails on curves.2.1

2.7.5.2 New Rail Profile Design

The new rail profiles are provided to the rail manufacturers. If the new rail
profiles were considerably incompatible with the wheel population in the sys-
tem, it would either cause poor initial wheel/rail contacts or require excessive
initial metal removal through grinding to correct the contact patterns.
In the past decade, the American Railway Engineering and Maintenance-of-
Way Association (AREMA) has revised the new rail profile for 57, 65.5, and
67.5 kg/m (136, 132, and 115 lb/yd) rails based on the assessment conducted
by TTCI. The revised new rail profiles have compatible contact features with the
population of wheels in North American freight service and only require minimal

2-48
metal removal if grinding is conducted on the new rail to remove the oxidation
layer. Figures 2-64 and 2-65 compare the new rail profiles before and after the
AREMA revision.2.33

Figure 2-64: New Rail Profiles before AREMA Revision

Figure 2-65: New Rail Profiles after AREMA Revision

Many routes in North America have curves with tight radii where excessive
wear and RCF are the concerns. Therefore, wheel/rail compatibility on curves
is a consideration and has led to modifications to the gage corner as shown
in figures 2-64 and 2-65. Also, the rail head radius has been revised from the
previous 254 mm (10 in.) to current 203.2 mm (8 in.). This is to form a more
rounded rail corner shape that provides a better match with the shape of the
worn wheel population.

2.7.5.3 Ground Rail Profile Design

Rail grinding is generally conducted to remove rail surface defects (including
corrugations) and/or to restore the transverse rail shape. Almost every railway
has its rail grinding templates designed based on its rail maintenance strate-
gies.2.1, 2.34

2.7.5.3.1 Ground Rail Profile for Tangent Track

As previously mentioned, a goal of designing a rail profile is to achieve low-
contact conicity. This can be achieved by producing a rail crown with large
radius [between 250 - 300 mm (9.8 - 11.8 in.)] and a lower gage corner that
leads to a strong, two-point contact on tangent when the wheel is flanging. This
limits the contact in the region of the wheel tread and rail crown and avoids large
rolling radius variations as the wheelset moves laterally on tangent track.
For a rail system with a high percentage of straight track with relatively
constant contact bands, the tangent ground template should especially consider
the conditions of worn wheels contacting worn rail. A small rail crown radius

2-49
has a tendency to cause hollow wheel tread. When the hollow wheels contact
slightly flattened, worn rail, a two-point contact with small contact areas is likely
to result, as illustrated in figure 2-66. This can lead to RCF on both wheels and
rails (figure 2-67).

Figure 2-66: Two Contact Bands with Both Having Small Contact Areas

Figure 2-67: Two RCF Bands

2.7.5.3.2 Rail Profile Grinding in Curves

Depending on the vehicle types and heavy haul railway system, the ground
high rail profiles may be designed separately for curves with large radii [> 800 m
(2,625 ft)] and small radius [< 800 m (2,625 ft)] to optimize contact. In general,
the tangent profile may apply to curves with radii larger than 2000 m (6,562 ft).
As discussed previously, near-conformal and conformal contact on the high
rail is desired. Therefore, design of high rail ground profile template(s) needs
to consider the wheel profile population in a system to ensure the ground high
rail profiles produce the desired conformal contact shape when contacting with
majority of wheels in the system.
Low rail profiles should generally have a center head radius of approximately
250 mm (9.8 in.) over a width of approximately 30 to 35 mm (1.2 to 1.4 in.) to
reduce contact stresses and to provide adequate rolling radius difference Δr for
curving.
Field side relief for both high and low rails needs to be carefully designed to
avoid contact toward the edge of the rails causing increased risk of rail roll and
high-contact stress, but also to avoid over cutting the field side that could result
in contact near the gage corner as shown in figure 2-53.

2-50
2.8 System Issues

2.8.1. Influence of Superelevation, Speed, and In-Train Forces on

Wheel/Rail Forces
Railway track is generally “banked” or superelevated in curves to counter
centripetal forces without appreciably transferring wheel loads between outer
and inner rails. In the limit, superelevation helps prevent overturning of the
vehicle. Inappropriate matching of superelevation to vehicle speed can adverse-
ly influence the curving performance of the vehicle. This, in turn, can lead to
differential settlement of the track between high and low rail and increase the
wear and stresses on both rail and wheel.

2.8.1.1 Balance Conditions

Curves are superelevated to balance the centripetal forces, associated with the
speed of the vehicle with gravitational forces related to the superelevation of the
track (figure 2-68).

Figure 2-68: Forces at Balance Speed in a Superelevated Curve

The ideal (or so-called balance speed) is given by:

V2 = (g s/2l) Rc (2.11)

where:
g = acceleration due to gravity
Rc = curve radius
2l = distance between wheel/rail contact points (track gage + one
rail head width)
V = speed

2-51
2.8.1.2 Influence of Cant Deficiency/Excess Superelevation on Verti-
cal Loads
Curving under unbalance conditions results in a transfer of vertical load
between high and low rails. Unbalance is often expressed in terms of the
amount of superelevation required to restore vertical load balance from current
superelevation and speed conditions. Example:
Figure 2-69 shows:
• No centripetal acceleration or associated forces are shown. It is assumed
that balance between the prevailing vehicle speed and the ideal
superelevation for that speed has been attained. su is the difference between
this ideal superelevation and the prevailing superelevation in the curve
(see section 2.8.1.5 and figure 2-71 for definition of su).
• Figure 2-69 shows s = su as an excess of superelevation.

Figure 2-69: Vertical Reaction Forces under Unbalance Conditions

For 2W (equal to the weight on one truck), acting at a distance d from the
centerline of the track, and for a center of gravity (CoG) height h, axle load W,
and zero lateral or roll deflection of the car body on the truck:

d = su h/2l

Consequently, reactions, RL and RR are:

R = W/2 (1 ± d/l), respectively (2.12)

2.8.1.3 Influence of Coupler Forces on Vertical Loads in Curves

Couplers become angled to the wagon’s longitudinal centerline in curves. This
angle is a function of wagon (car) geometry and coupler design and that of the
adjacent car or locomotive.
Longitudinal train forces act at a height of the coupler. The lateral component
of the coupler force, when the coupler is angled in a curve, results in wheel load
transfer between high and low rail (figure 2-70).

2-52
 Figure 2-70: Lateral Component of Coupler Force

2.8.1.4 Influence of Superelevation Unbalance and In-Train Forces

on Vertical Loads
A combination of excess superelevation and tensile forces in the couplers of a
car (generally in the leading cars of a train) will result in excess load on the low
rail. A combination of insufficient superelevation and compressive forces in the
couplers (generally the trailing cars in a train with locomotives in the rear) will
result in excess load on the high rail.
These forces are readily calculated given speed, car load, height of car’s center
of gravity above rail and the height of the coupler above rail, as well as car and
wheelset geometry. Care should be taken in establishing the role of the vertical
(roll) and lateral deflection of the car body under unbalance conditions as this
can be considerable.

2.8.1.5 Influence of Superelevation Unbalance and In-Train Forces

on Lateral Loads
Curve negotiation with cant unbalance will also result in increased lateral loads
on both rails. Figure 2-71 shows that lateral loads L must be generated at the
wheel/rail interface to counter the gravitational component of the vertical load.
If 2W equals the weight of half a car, the lateral wheel/rail contact forces
required to be reacted by one truck L:

L= 2W su/2l (2.13)

Considering a car weight of 286,000 pounds, L = 2,403 pounds per inch of cant
unbalance (or for su = 1). This implies that if the lateral force were to be “shared”
equally between the four wheels of the truck, each wheel would experience a
lateral load of 2,403/4 = 601 pounds per inch of cant unbalance.

2-53
 Figure 2-71: Lateral Forces on a Bogie/Car Due to
Superelevation Unbalance
The lateral component of the coupler force must be similarly “shared” between
wheel/rail contact points.
Explained in section 2.4, railway wheelsets produce lateral forces at the wheel/
rail interface by aligning at an angle of attack α to the track (figure 2-72a).
Similarly, wheelsets in bogies produce lateral forces at the wheel/rail interface
in reaction to lateral forces at the center plate by aligning at an angle of attack α
to the track (figure 2-72b). Consequently, bogies align in curves with increased
angles of attack under excess cant conditions and with reduced angles of attack
under conditions of cant deficiency.

(a) Single Wheelset (b) Bogie

(c) Influence of Flange Forces

Figure 2-72: Lateral Forces on a Wheelset/Bogie

2-54
 Generally, because of the geometry of the curve and of the bogie, the angle of
attack of the lead wheelset is greater than that of the trail wheelset with the angle
of attack of the trail wheelset being approximately zero. Consequently, most of
the lateral force to counter cant unbalance is generated by the lead wheelset.
High rail flange contact creates flange forces on the lead wheelset that act in a
similar sense to forces due to curving with excess cant (figure 2-72c). This can
increase the lateral forces on the lead wheelset when curving with flange contact
under conditions of excess cant.

2.8.2 Strategies for Designing Curves and Superelevation and

Managing In-Train Forces
The conclusion to be drawn from the above descriptions of the influence
of speed, superelevation, and in-train forces is that curving at balance speed
with zero coupler forces provides the “best” solution. Obviously, this is not a
practical solution under practical operating conditions as train speeds can vary
(typically between empty and loaded trains on grades) and in-train forces are
always present. However, generally, balance speed should be sought for the
greatest tonnage. Figure 2-73 shows a typical distribution of MGT vs. speed
in a specific curve on a (predominantly) heavy haul operation. Obviously, the
ideal design would be for balance conditions at between 16 and 24 kph (10 and
15 mph). The question remains: How to cater to trains running, particularly at
speeds higher than 24 kph (15 mph) and as high as 72 kph (45 mph)?

Figure 2-73: Tonnage Distribution vs.

Figure 2-74: Superelevation vs.
Speed (4.5 degree curve) Speed (4.5 degree curve)
A review of the conclusions drawn in sections 2.8.1.4 and 2.8.1.5 suggest that
curving with superelevation deficiency is probably a better option than that of
curving with excess superelevation.
Although curving with superelevation deficiency, it increases the load on the
high rail, it also:
• Reduces the angle of attack of the lead wheelset, tending to reduce lateral
forces
• Reduces the lead axle high rail L/V ratio [higher vertical load (V) with a
lower lateral load (L)]

2-55
 This would reduce the propensity for flange climb and suggests that high rail
wear and flange wear would not be increased and may be decreased.
Superelevation deficiency is generally limited to 76 mm (3 in.) in North
America with some exceptions.
Curving with excess superelevation increases the vertical and lateral loads
on lead axle, low rail contact. This is understood to be a leading cause for
high-impact wheels, particularly in North America and is also understood to be
a driver of low rail head damage through rolling contact fatigue (RCF), wear,
material flow, and crack formation.
Excess superelevation is generally limited to a static wheel load loss of 40% for
a wagon (car) at zero speed in a curve.
Given the foregoing, a design can often be obtained for balance for the
preponderance of traffic at balance speed. Figure 2-74 shows a graph of balance
speed vs. superelevation with superelevation design options to remain within
a 76 mm (3 in.) superelevation deficiency limit. It references figure 2-73 and
suggests that zero superelevation would provide balance for 83% of traffic with
a speed limit of 50 kph (31 mph) [76 mm (3 in.) superelevation deficiency];
however, if a 25.4 mm (1 in.) superelevation were provided, the speed limit would
be 56 kph (35 mph). The current design is for 89 mm (3.5 in.) superelevation
providing a maximum speed in excess of 72 kph (45 mph) and catering to 8%
of the lightly loaded traffic, but in the process introducing unnecessarily high
vertical and lateral loads on the low rail.
To reduce the effect of traction forces on vertical and lateral loads on the
contact patch, distributed power (DP) is recommended to limit in-train forces
on grades.

2.8.3 The Influence of Curve Radius

In general, when a wagon bogie with a high stiffness negotiates a curve with a
small radius, the outer wheel flange of the front wheelset will contact with the rail
head (figure 2-75), while the position of the rear wheelset in the curve depends
on the running speed. A low speed may result in the flange contact of the inner
wheel on the trailing axle, while a moderate speed will never generate flange
contact of both the inner and the outer wheels. However, as the speed reaches
a certain value, contact of the outer wheel flange of the trailing axle will occur.

Figure 2-75: Bogie Position in Curve

2-56
 For the front wheelset, there is an angle of attack between the wheel plate and
the tangential line of the outer rail in the curve at the wheelset location. To a
certain extent, the size of the attack angle influences the wheel/rail lateral
dynamic interaction. A large angle of attack directly threatens the running
safety. Figure 2-76 shows the angles of attack on two curves with different radii.
Note that the angle of attack α1 in the curve with small radius is larger than that
with the big radius.

Figure 2-76: Attack Angles of Wheelset on Curves with Different Radii

Another important issue is to determine what curve conditions will not produce
flange contact. Figures 2-77 and 2-78 show the relationship between curve, truck
wheelbase, conicity, and wheelset lateral displacement.2.1 Typically, bogies on
standard gage, with wheelbases of approximately 1.8 m (5.9 ft) may negotiate
curves with radii between 1500 and 2000 m (4,922 and 6,562 ft) without flange
contact. Under these conditions, the lateral and longitudinal creepages are low
when the bogies are accurately aligned, thus minimal rail and track damage is
experienced. Side and crown wear are minimal, with a degree of material flow
to the field side of both high and low rails after about 200 MGT. This may be
corrected by grinding.
In sharp curves, below a radius of 1500 m (4,922 ft), with low wheelset conicity
and reduced gage clearance, at large track discontinuities, or when misaligned
bogies are present in the vehicle fleet, flange contact takes place and the bogie
takes an attitude as figure 2-76 shows.

2-57
 Figure 2-77: Lateral Wheelset Displacement vs. Curve Radius

Figure 2-78: Lateral Wheelset Displacement vs.

Bogie Wheelbase

2.8.4 The Influence of Fasteners

The main functions of fasteners are to secure rails to sleepers and reduce rail
vibration. Their influence should be considered together with other system
issues:
• With regard to running stability, increasing the fastener lateral stiffness
will strengthen the rigidity of the track structure in favor of improving the
wagon critical speed.
• In sharp curves, the main dynamic problem is the large wheel/rail lateral
force. From the perspective of track structure damage, the lateral
stiffness should be increased to avoid large rail deformation (figure 2-79).
But the disadvantage is that lateral wheel/rail dynamic interaction and
vehicle running safety may be adversely affected.

2-58
 • At welded rail joints, the wheel/rail impact loads occur constantly with a
high frequency force P1 and a low frequency force P2.2.41 Low vertical
stiffness of the fastener is helpful to alleviate the impact load
(figure 2-80).2.42

Figure 2-79: Effect of Fastener Lateral Stiffness on Lateral Force

and Track Gage

Figure 2-80: Effect of Fastener Vertical Stiffness on

Wheel/Rail Vertical Impact Forces

2.8.5 The Influence of Track Structure Parameters – The Chinese

Experience
The track structure is a complicated system in which each component and
parameter has its particular role in the dynamic performance. In this section,
emphasis is placed on the influence of asymmetries, fasteners, track gage, and
curve radius on the dynamic behavior of freight wagons.

2-59
2.8.5.1 Asymmetries of Wheel/Rail Contacts and Forces on Curved
Track
On curved track, the wheelset usually has a lateral displacement toward the
outer rail, especially for the front wheelset, which leads to asymmetries of wheel/
rail contact. In this case, the contact parameters of the outer and inner wheels,
such as the contact positions, rolling radius, and contact angle are different from
those of the straight track case (figure 2-81).

Figure 2-81: Wheel/Rail Contact Geometry in Curves

Typical wheel/rail contact distributions on the surfaces of the standard rail
profile are shown in figure 2-82, where the lateral displacement of the wheelset
is from 0 to 15 mm (0 to 0.6 in.) toward the outer rail. The contact positions of
the outer rail mainly occur in the regions near the rail side and the rail crown.
The points have a discontinuity in the contact region between the wheel flange
root and the rail gage corner, which usually leads to an incremental load on
the rail side and thus the increase of the rail side wear. For the inner rail, the
contact points distribute uniformly in the local region of the rail top. Obvious-
ly, the asymmetries of wheel/rail contacts in curves result in different failure
characteristics for the outer rail and the inner rail, as shown in figure 2-83 where
the shiny regions reflect the wheel/rail contact bands after rail wear. In sharp
curves, the common defect in the inner rail is spalling, which mainly occurs in
the center region of the rail top. For the outer rail, severe side wear occurs and
the rail profile nearly conforms to the flange shape.

Figure 2-82: Wheel/Rail Contact Positions of Chinese LM Wheel Profile with Stan-
dard Rail Profile

2-60
 (a) Outer rail (b) Inner rail
Figure 2-83: Asymmetric Profiles of Outer and Inner Rails in a Sharp Curve
Based on wheel/rail contact theory, the asymmetric wheel/rail contact geom-
etries will obviously change the force conditions of two wheels fixed on a single
axle. Actually, the forces acting on the vehicle in curves are extremely complex,
such as the centrifugal force, wind force, wheel/rail normal force, wheel/rail
tangential force (including the longitudinal and lateral creep forces), and gravity
[figure 2-84(a)]. The primary forces ultimately on the rail can be projected to
be a wheel/rail vertical force and a lateral force [figure 2-84(b)]. They are two
important indicators for evaluating the dynamic behavior of the vehicle and the
track. The magnitudes of these two forces are significantly influenced by many
factors such as the wheel/rail contact condition, vehicle running speed, and
external forces.
In general, if the vehicle in a curve operates with a deficient superelevation, the
car body rolls and moves laterally toward the outer rail, and vehicle load transfer
occurs. In this case, the outer rail will bear a larger vertical force than the in-
ner rail. Conversely, a surplus superelevation generally leads to a vertical load
decrease on the outer rail and a load increase on the inner rail. But for the wheel/
rail lateral force, the asymmetries are closely related to the steering ability of
the wheelset. Basically, on the same curved track, the wheel/rail lateral force
of the front wheelset in the wagon bogie increases as the running speed rises.
Under certain conditions of deficient superelevation and wagon parameters, the
rear wheelset of a wagon bogie has a tendency to approach the outer rail, which
could share and reduce the lateral load on the front wheelset. In other words, the
deficient superelevation has a positive influence on the curving performance of
the wagon bogie in this case.

2-61
 Figure 2-84: Schematic of Forces Acting on a Vehicle
in a Curve
The asymmetries of the wheel/rail forces on the curved track greatly depend
on the track structure parameters, vehicle performance, and wheel/rail con-
tact status. It should be analyzed under the actual track and vehicle operating
conditions.

2.8.5.2 Motion of Rails in Curve

Generally, the rails are connected to the sleeper with the fasteners (fig-
ure 2-85). The elastic fastener and rubber pad in a fastening system are the
main components that allow the rail motion. At the same time, they can also
attenuate the vibration. Under the action of the wheel/rail forces, the elasticity of
the fastening and the track structure provides the possibility for rail deformation.

Figure 2-85: Typical Rail Fastening System

On tangent track, the wheel load produced by gravity is the main
component of the wheel/rail vertical force. Therefore, the downward motions in
the vertical direction of the rails are the common phenomena (figure 2-86). But in
curves, vertical deformations of the outer and inner rails are unequal due to their
different wheel loads. Wheel load reduction will decrease rail deformation while
an increase of wheel load enlarges it.
The lateral motion at the rail torsional center is mainly affected by the wheel/
rail lateral force (figure 2-86). For the wagon bogie in a curve, large wheel/rail
lateral forces usually occur due to the steering action of the front wheelset.

2-62
 The rail rotational motion is determined by the resultant moment M, calculated
by:

M = Qhr + FLa - Pe + (FV1 -FV1)b (2.14)

There are three different cases:

• Case 1: M=0, no rail rotation occurs.
• Case 2: M>0, the rail rotates clockwise.
• Case 3: M<0, the rail rotates counterclockwise.

Figure 2-86: Rail Motion and Forces

In curves, severe wheel/rail lateral interaction commonly occurs in the outer
rail. Track gage is widened owing to rail deformation, as shown in figure 2-87.
A large track gage enlargement could cause the wheels on the inner rail to
drop and track components would be damaged. If the deformation of the rail is
controlled within an allowable range, long-term service security of the track
structure can be guaranteed. Once overload occurs, the track structure may
be damaged, and this has a harmful influence on the running safety of freight
wagons.

Figure 2-87: Track Gage Enlargement (Widening)

It is necessary to evaluate the influence of rail deformation on the running
safety of freight wagons. For dynamic analysis, the vehicle/track coupled
dynamics provide an effective method to deal with such problems.

2-63
3 N 2.8.5.3 The Influence
P₀,max = 32 -of
N Track Gage
2 Ae √ 1 - ( aξ )² - ( ηb )² , Ae
A lateral clearance is kept
Δ tan(δ) - μ between the rail gage corner and wheel flange
(figure
0.3 ∙ p₀ at z = 0.78a.
2-88). When (
(32) =a max
wheelset
1 + μ *
)
tan(δ)rolls on a straight track, it will move laterally
by a distance δ to “find” equal rolling diameters. If the gage clearance Δmax is
insufficient to accommodate the distance δ, flange contact will occur, resulting
in flange and gage corner wear. The lr0
value of Δvmax and the track gage will change
Rw δ when ythe l
r0rails and λ
wheels = 2πare worn f = 0the gage is widened. From the
or after
= γ
R w - RR perspectivec of wheel/rail contact geometry,λ their influence on the wheel/
R γ
rail interaction can be illustrated qualitatively by studying the conicity
characteristic. The effective conicity γ of a wheelset is defined as the wheel
Effective Conicity (λ) = RRD
[
Δr = r0 rolling
]
R + l - 1 radius differential (RRD) divided by twice the lateral motion caused by
2y
R - l radii difference. This relationship is given in equation 2.15. In most cases, the
the
conicity increases with increasing RRD.

γ = rolling radius differential

(2.15)
2y

y R R δ2l λ π w e
R

Figure 2-88: Track Gage Clearance

By moving the left rail in figure 2-89, different track gages can be achieved. The
contact points with respect to the enlarged, standard, and narrowed track gages
are indicated by A, B, and C, respectively.
When the track gage is narrowed, the wheel/rail contact point on the left wheel
will be changed from point B to C. Due to the effect of the coned wheel profile,
the rolling radius increases from rLB to rLC, which enlarges the RRD value. In
curves, it is helpful to utilize the creep forces to steer near the radial position
without wheel flange contact, alleviating the amplitude of the wheelset lateral
motion. However, the hunting motion stability will be affected.
Conversely, if the track gage is enlarged, the left wheel contact point moves
from point B to A with a reduction of rolling radius from rLB to rLA. The RRD value
is decreased in this case. As a result, the wheelset creep steering ability in curves
will be limited and flange contact may frequently occur. Certainly, there will be
an increased tendency for larger wheelset lateral motion amplitudes. However,
the smaller RRD is able to improve the hunting stability of freight wagons on
tangent track.

2-64
 Figure 2-89: Wheel/Rail Contact with Different Track Gages
From the aspect of geometric curving of freight wagons, gage clearance must
be maintained so that the wheelset can pass through a curve smoothly. The
influence of gage clearance on the running resistance, energy loss, and wear of
wheels and rails, etc., should always be considered.

2-65
2-66
References

2.1. Guidelines to Best Practices for Heavy Haul Railway Operations: Wheel and Rail
Interface Issues, First Edition, International Heavy Haul Association, Virginia
Beach, VA, 2001.
2.2. Tournay, H., “Wheel Rail Interaction from a Track and Vehicle Design
Perspective,” Procedings of Conference on Wheel/Rail Interface, IHHA, Moscow,
June 14-17, 1999.
2.3. Redtenbacher, F.J., “Die Gesetze des Locomotiv-Baues,” Verlag von Friedrich
Bassermann, Mannheim, p. 22, 1855.
2.4. Dendy Marshall, C.F., “A History of British Railways down to the Year 1830,”
Oxford University Press, London, pp. 147-148, 1938.
2.5. Klingel, J., “Über den Lauf von Eisenbahnwagen auf Gerader Bahn,” Organ
Fortschritte des Eisenbahnwesens, Neue Folge 20, pp. 113-123, 1883.
2.6. Andersson, E., Berg, M., and Stichel, S., “Rail Vehicle Dynamics,” KTH Railway
Group Royal Institute of Technology, Stockholm, Sweden, 2013.
2.7. Knothe, K. and Stichel, S., Schienenfahrzeugdynamik, Springer-Verlag Berlin
Heidelberg, 2003.
2.8. Scheffel, H., Tournay, H.M., and Fröhling, R.D., “The Evolution of the Three-Piece
Freight Car Bogie to Meet Changing Demands in Heavy Haul Railroads in South
Africa,” Proceedings of the Conference on Freight Car Trucks/Bogies organized by the
International Heavy Haul Association, Montreal, Canada, June 1996.
2.9. Wickens, A.H., “The Dynamic Stability of Railway Vehicle Wheelsets and Bogies
having Profiled Wheels,” International Journal of Solids and Structures, vol. 1,
pp. 319-341, 1965.
2.10. Joly, R., “Study of the Transverse Stability of a Railway Vehicle Running at High
Speed,” Rail International, vol. 3, no. 2, pp. 83-117, 1972.
2.11. Scheffel, H., “The Influence of the Suspension on the Hunting Stability of Railway
Vehicles,” Rail International, pp. 662-696, August 1979.
2.12. Scheffel, H., “Die Konstruktion der Kreuzanker-Drehgestelle der Südafrikanischen
Eisenbahnen,” ZEV-Glasers Annalen 110, Nr. 6/7, Juni/Juli 1986.
2.13. Scheffel, H. and Tournay, H.M., “An Analysis of Steady State Deviations of
Wheelsets from the Pure Rolling Position Caused by Curved Track and Bogie
Inaccuracies,” Proceedings of the 8th IAVSD Symposium, MIT, Cambridge, 1983.
2.14. Scheffel, H., Fröhling, R.D., and Heyns, P.S., “Curving and Stability Analysis of Self
Steering Bogies having a Variable Yaw Constraint,” Proceedings of the 13th IAVSD
Symposium, Chengdu, People’s Republic of China, 1993, Supplement to Vehicle
System Dynamics, volume 23, pp. 425-436, 1994.
2.15. Scheffel, H., “A New Design Approach for Railway Vehicle Suspension,” Rail
International, pp. 638-651, October 1974.
2.16. Scheffel, H., “Self-Steering Wheelset Will Reduce Wear and Permit Higher Speeds,”
Railway Gazette International, pp. 453-456, December 1976.
2.17. Scheffel, H. and Tournay, H.M., “The Development of an Optimal Profile for
Self-Steering Trucks under Heavy Axle Load Conditions,” Rail Transport Division
of the American Society of Mechanical Engineers (ASME), Paper 80-WA/RT-5,
Winter Annual Meeting, Chicago, Illinois, U.S., pp. 1-12, November 1980.

2-67
2.18. Stanley, I.N., Philadelphia, PA, “Improvement in Railroad Cars for Turning
Strong Curves, and for Changing from One Track to Another Without the Use of
Switches,” patented March 12, 1840.
2.19. Smith, R.E. and Anderson, R.J., “Characteristics of Steered Railway Trucks,” IAVSD
Journal, vol. 17, pp. 1-36, 1988.
2.20. Anderson, R.J. and Fortin, C., “Low Conicity Instabilities in Forced-Steering
Railway Vehicles,” Proceedings of the 10th IAVSD Symposium, Prague,
August 24-28, 1987.
2.21. Smith, R.E., “Steering Rail Vehicle Axles – A Historical Review,” Proceedings of the
Canadian Engineering Centennial Convention, Montreal, May 1987.
2.22. Tournay, H., “Rail/Wheel Interaction from a Track and Vehicle Design
Perspective,” IHHA STS Conference, Moscow, Russia, June 14-17, 1999.
2.23. Elkins, J.A. and Weinstock, R., “The Effect of Two Point Contact on the
Curving Behavior of Railroad Vehicles,” AMSE 82-WA/DSC-13, American Society of
Mechanical Engineers, New York, NY, 1982.
2.24. Mace, S. et al., “Effects of Wheel-Rail Contact Geometry on Wheelset Steering
Forces,” Wear, vol. 191, issues 1-2, pp. 204-209, 1996.
2.25. Wu, H. and Kerchof, B., “Management of Wheel/Rail Interface to Prevent Rail
Rollover Derailments,” 2013 IHHA conference, to be published in JRRT.
2.26. Wu, H., Woody, S., and Blank, B., “Optimization of Rail Grinding on a North
American Heavy Haul Railroad Line,” Proceedings of 2005 International Heavy Haul
Association Conference.
2.27. Sawley, K. and Wu, H., “The Formation of Hollow-Worn Wheels and their Effect on
Wheel/Rail Interaction,” Wear, 258 (7-8), pp. 1179-1186.
2.28. Tournay, H., Wu, H., and Guins, T., “The Influence of Hollow-Worn Wheels on
the Incidence and Costs of Derailments,” Research Report R-975, Association of
American Railroads, Transportation Technology Center, Inc., Pueblo, CO, U.S.,
February 2004.
2.29. Wu, H. and Madrill, B., “Effect and Formation of Asymmetrically Worn Wheels,”
Technology Digest TD-08-021, Association of American Railroads, Transportation
Technology Center, Inc., Pueblo, CO, U.S., 2008.
2.30. Blader, F.B., “A Review of Literature and Methodologies in the Study of
Derailments Caused by Excessive Forces at the Wheel/Rail Interface,” Research
Report R-717, Association of American Railroads, Transportation Technology
Center, Inc., Pueblo, CO, U.S., December 1990.
2.31. Leary, J., “Final Report on the Development of an Alternative AAR Interchange
Wheel Profile,” Report No. R-706, Association of American Railroads, November
1988.
2.32. Wu, H., “Effects of Wheel and Rail Profiles on Vehicle Performance,” Proc.
International Association of Vehicle System Dynamics IAVSD Conference, Milano,
Italy, August 2005.
2.33. American Railway Engineering and Maintenance-of-Way Association Manual, 2005,
2006.
2.34. Sroba, P. and Roney, M. et al., “The Evolution of Rail Grinding on Canadian Pacific
Railway to Address Deep Seated Shell in 100% Effective Lubrication Territory,”
World Congress on Railway Research, Montreal, Quebec, Canada, 2006.

2-68
2.35. Zhai, W.M., Wang, K.Y., and Cai, C.B., “Fundamentals of Vehicle-Track Coupled
Dynamics,” Vehicle System Dynamics, 47(11), pp. 1349-1376, 2009.
2.36. Zhai, W.M., Cai, C.B., and Guo, S.Z., “Coupling Model of Vertical and Lateral
Vehicle/Track Interactions,” Vehicle System Dynamics, 26(1), pp. 61-79, 1996.
2.37. Zhai, W.M. and Wang, K.Y., “Lateral Interactions of Trains and Tracks on
Small-Radius Curves: Simulation and Experiment,” Vehicle System Dynamics,
44 (Suppl.), pp. 520-530, 2006.
2.38. Zhai, W.M. and Wang, K.Y., “Lateral Hunting Stability of Railway Vehicles Running
on Elastic Track Structures,” Journal of Computational and Nonlinear Dynamics,
ASME, 5(4), pp. 041009-1-9, 2010.
2.39. Andersson, E., Berg, M., and Stichel, S., “Rail Vehicle Dynamics,” KTH Railway
Group Royal Institute of Technology, Stockholm, Sweden, 2013.
2.40. Knothe, K. and Stichel, S., Schienenfahrzeugdynamik, Springer-Verlag Berlin
Heidelberg, 2003.
2.41. True, H., “On the Theory of Non-Linear Dynamics and its Applications in Vehicle
System Dynamics,” Veh. Syst. Dyn., 31(5), pp. 393-422, 1999.
2.42. Polach, O., “On Non-Linear Methods of Bogie Stability Assessment using
Computer Simulations,” Proceedings of the Institution of Mechanical Engineers, Part
F: Journal of Rail and Rapid Transit, 220, pp. 13-27, 2006.
2.43. Jenkins, H.H. et al., “The Effect of Track and Vehicle Parameters on Wheel/Rail
Vertical Dynamic Forces,” Railway Engineering Journal, 3(1): 2-16, 1974.

2-69
2-70
Contributors to Chapter 2

Harr y Tournay
Mr. Tournay is Senior Scientist at Transportation
Technology Center, Inc., and for the past 12 years has
conducted mechanical and vehicle/track research under
the Association of American Railroads strategic research
program. Projects include the development of an improved
freight bogie, investigating the root causes of (rail and
wheel) rolling contact fatigue, investigating the effects of
superelevation on freight vehicle/track interaction, brake
technology, and wayside detection.

This follows a 29-year railroad career culminating as Chief Engineer, R&D at

Spoornet (South Africa); responsible for mechanical, track, electric power, and
signaling. He was a director of the International Heavy Haul Association and is a
member of the IHHA Heavy Haul of Fame. During his career in South Africa, he
was involved with vehicle and track design, maintenance, and strategic planning,
particularly in heavy haul operations.

A mechanical engineer by education and training, he worked in research and

development with a focus on wheel/rail interaction as well as the design and
construction of coal and other wagons and of Scheffel steering bogies for freight
and passenger operation as well as for locomotives. His particular focus was
vehicle curving behavior and wheel and rail wear mechanisms.

Robert Fröhling
Dr. Fröhling has more than 30 years of railway experience with extensive
experience in rail vehicle system dynamics, vehicle/track interaction, wheel/rail
interaction, bogie technology, and structural mechanics. He has published and/
or presented 47 international papers on these subjects. He is Principal Engineer
within the Technology Management Department of Transnet Freight Rail and
he has responsibility for the following core Mechanical Railway Technologies:
Railway Vehicle System Dynamics, Wheel/Rail Interaction, Bogie Technology,
Structural Mechanics, as well as Locomotive and Wagon Mechanical Design
Integrity.

2-71
Roy Smith
Mr. Smith, P.Eng., M.Eng., B.Sc. has been the president of RESCO Engineer-
ing since he formed the company in 1989. Prior to that, between 1975 and 1989,
he worked on rail car truck design at UTDC (now part of Bombardier). He
graduated from the University of Birmingham (UK) in 1963 and from the
University of Toronto in 1970. The author of many technical papers on the
subject of truck dynamics and curving, Mr. Smith also has over twenty patents
to his name. He continues to be active in the development of technology for
the improvement of railway economics and safety. He was born in England and
emigrated to Canada in 1965, where he has lived ever since.

Sebastian Stichel
Prof. Stichel holds a Ph.D. (1996) in Vehicle Dynamics from TU Berlin.
He has been a professor at the Royal Institute of Technology in Stockholm
since 2010. He is director of the KTH Railway Group, vice chairman of the
Department of Aeronautical and Vehicle Engineering, and director of the mas-
ter’s program in Vehicle Engineering at KTH. From 2000 to 2010, Dr. Stichel
was employed at Bombardier Transportation in Sweden where he headed a
Vehicle Dynamics department with employees in Sweden, Germany, UK, and
France since 2003. His primary research interest is in dynamic vehicle/track
interaction using multi-body simulation including improved ride comfort and
reduced wheel and track damage.

Huimin Wu
Dr. Wu is a Principal Investigator at Transportation Technology Center, Inc.,
Pueblo, Colorado, U.S. She holds a Ph.D. degree in Mechanical Engineering
from the Illinois Institute of Technology, Chicago, U.S. Dr. Wu has more than
20 years of experience in railroad research areas of vehicle dynamics, wheel/
rail interaction, wheel/rail profile design, rail grinding, automated wheel/rail
contact inspection systems, and rail wear and RCF management model.

Wanming Zhai
Prof. Zhai is a Professor at Southwest Jiaotong University, a member of the
Chinese Academy of Sciences and the Editor-in-Chief of International
Journal of Rail Transportation. He has developed a model of vehicle-track coupled
dynamics and a method to analyze and evaluate high-speed train-track-bridge
dynamic interaction for Chinese high-speed operations. He graduated from
Southwest Jiaotong University in 1985, receiving a Ph.D. in Railway Vehicle
Engineering in 1992.

2-72