Considerations for Mechanistic Design of Concrete Sleepers and

Elastic Fastening Systems in North America

B.J. Van Dyk, J.R. Edwards, C.J. Ruppert, Jr., C.P.L. Barkan
Rail Transportation & Engineering Center (RailTEC),
University of Illinois at Urbana-Champaign (UIUC),
Urbana, IL, USA

Abstract

A sustained increase in heavy axle loads and cumulative freight tonnages, as well as increased
interest in high speed passenger rail development, has placed an increasing demand on railway
infrastructure and its components. One of the most critical areas of the infrastructure in need of
further research and analysis is the concrete sleeper and elastic fastening system used in heavy
haul and shared infrastructure applications. A limited understanding of the complex loading
conditions affecting the concrete sleeper and elastic fastening system components led to a
design process based primarily on practical experience and previous techniques, which fails to
include key variables that relate to actual field loading conditions. This process, which is typically
driven by production and installation economics, has generated components that do not achieve
their design life. While initially functional, they ultimately require more frequent maintenance or
fail prematurely, causing track outages, reduced capacity, and added cost. To address this
challenge, the University of Illinois at Urbana-Champaign (UIUC) is analyzing and comparing the
existing heavy haul and shared infrastructure loading environment to current recommended
design practices within the international railway community. This paper discusses the use of field
experimental data, as well as complete sleeper and fastening system analytical modeling, and
how they can be used to improve the current understanding of the loading demands on each
component within the system. Results from these tests and modeling efforts, along with
probabilistic considerations with respect to dynamic and impact loading, will contribute to a
greater understanding of the loading regime present in the concrete sleeper and elastic fastening
system in heavy haul and shared infrastructure applications. Ultimately, this improved
understanding will provide a basis for a mechanistic design process, contributing to improved
recommended practices for concrete sleeper and fastening system design and improving safety,
reliability, and rail capacity.

1. Introduction

Historically, the North American sleeper (or crosstie) and fastening system have been designed
through a process that is generally based on practical experience, without a clear understanding
of failure mechanisms, their causes, and the loading environment. This design methodology has
led to performance challenges and service failures that cannot be adequately explained or
predicted. Without a clear framework for the design and expected performance of concrete
sleepers and fastening systems, inefficiencies in component design and maintenance may exist,
with a resulting negative impact on the economics of concrete sleepers and fastening systems.
Improvements in the design of these systems will provide a more robust railway superstructure,
where the loading environment is more fully understood, failures are reduced, and the possibility
of predicting performance and wear rates exists.

The North American loading environment differs from that throughout much of the rest of the
world, due to the prominence of rail freight transport and sharing of infrastructure between freight
and passenger traffic. This paper will investigate the particular loading conditions found in North
America and draw comparisons between the varied international design considerations where the
loading environment may be different.

2. Current Recommended Design Practices

Internationally, there are many unique design methodologies for the manufacture of concrete
sleepers and fastening systems. Many countries have their own version of design standards or
recommended practices that railways and manufacturers follow to varying degrees. This paper
will briefly discuss the similarities and differences in design methodologies found in North
America, Europe, Australia, and Japan.

The American Railroad Engineering and Maintenance-of-way Association’s (AREMA) Manual for
Railway Engineering is the primary source of guidance for the design and construction of North
American railway infrastructure. It is a set of recommended practices, and is typically modified by
individual railways to meet their specific loading or performance objectives [1]. Chapter 30 of the
Manual for Railway Engineering provides guidance for sleepers, and Part 4 of that chapter
focuses on concrete sleepers. While this section of the AREMA manual offers helpful information
for railways and sleeper manufacturers, there are opportunities for improvement, particularly in
terms of the sleeper design process (hereafter referred to as the “AREMA Method”).

One opportunity for improvement of AREMA Chapter 30 is the consideration of component

interactions and system performance. In the 2012 International Concrete Sleeper and Fastening
System Survey conducted by UIUC, fastening system manufacturers indicated that component
and system interaction plays a large role in their design [2]. This concept should be included in
the development of improved design recommendations for concrete sleepers and fastening
systems.

There are two design parameters used in the AREMA Method for determining concrete sleeper
geometric and strength characteristics: allowable ballast pressure and flexural performance. In
determining the allowable ballast pressure, the AREMA Method considers sleeper spacing
(leading to the determination of a load distribution factor), wheel load, an assumed impact factor,
and sleeper bearing area. Another portion of the AREMA Method for concrete sleepers contains
the flexural performance requirements. These requirements consider sleeper length, sleeper
spacing, speed, and tonnage to determine the positive and negative design bending moments at
the center of the sleeper and at the rail seat. Some consideration was given to impact or dynamic
factors and axle loads in the fabrication of the method, but they are not explicitly used in the
recommended practices for flexural design. Therefore, the flexural design of a concrete sleeper
as found in AREMA Chapter 30 does not consider many important design criteria, such as track
geometry (e.g. curvature and grade), design life, or impact factors and axle loads that reflect the
intended loading environment.

Because it is typically the primary design criteria of concrete sleepers, the authors have reviewed
the design bending moment design methodologies in multiple standards and recommended
practices. Within each methodology, different design principles are considered and used.

The European Standard offers recommendations for the design of concrete sleepers, and, like
the AREMA Method, its primary focus is the design bending moment. However, EN 13230 states
that the specific design method is the responsibility of the purchaser, considering static and
dynamic wheel loads, design and maintenance of the track (including longitudinal distribution of
wheel loads), climatic conditions, magnitude of prestressing force, strength of concrete, and
particular, non-standard designs [3].

The Australian Standard calculates positive and negative rail seat and center bending moments
using sleeper spacing, static wheel load, track modulus, rail modulus, rail second moment of
area, quasi-static and dynamic design load factor, sleeper length, gauge, and support conditions
[4]. An intermediate step to this process incorporates Talbot’s method for determining rail seat
loads [5]. The standard also explicitly states that sleeper sections need not be checked for
stresses other than flexural stresses [4].

The Japanese Industrial Standard (JIS) simply provides “bending forces” that must be exceeded
during testing of concrete sleepers [6]. The design methodology is not explicitly provided in the
JIS, and is therefore determined by the manufacturer, as long as it meets the performance criteria
as stated in the JIS.

After reviewing the above international design methodologies, it is evident that the concrete
sleeper design process is not uniform throughout the international railway community. There are
many criteria to be considered from design recommendations and best practices worldwide.
These principles can be applied to the development of an approach that is centered on science
and materials properties to govern the design of concrete sleepers and fastening systems in
North America. However, the operating environment in North America, which is often different
than that found elsewhere in the world, must be better understood before mechanistic design
recommendations can fully be developed and placed into practice.

3. Principles of Mechanistic Design

The mechanistic design process is one derived from analytical and scientific principles,
considering field loading conditions and performance requirements. Mechanistic design has been
used in other disciplines, such as the design of rigid and flexible highway pavements using
particular input values, performance analyses, and alternative evaluations [7].

Historically, North American concrete sleepers and elastic fastening systems have been designed
through a design process that does not include all of the critical variables relating to actual field
loading conditions. A lack of understanding regarding the complex loading conditions of the
system has led to a design methodology driven by production and installation economics, where
very high priority is placed on manufacturing and installation efficiency. Oftentimes, this process
is not directly based on actual performance of the sleeper and fastening system or a thorough
understanding of the demands on each component.

Therefore, UIUC is developing a mechanistic design process that uses the existing loading
environment and sleeper and fastening system components. This exercise will create an
improved understanding of failure causes and effects on performance. Design would typically be
directed toward a specific failure mode (often grouped into one of three categories; support,
stability, or isolation failure [8]), creating predictable wear and fatigue rates and leading to repair
cycles that coincide with other planned maintenance intervals. This improved design procedure
will increase production and operational efficiency while reducing unscheduled maintenance and
track outages.
3.1. Shared Use Loading Environment in North America

The railway operating environment in North America is different than much of the rest of the
world. As enthusiasm for higher-speed intercity passenger service grows, some “incremental”
systems are developing that require passenger and freight traffic to share the same infrastructure.
Shared railway infrastructure provides an effective method for providing an incremental approach
to higher-speed passenger transportation, and reduces the first cost associated with opening a
new system. One of the many challenges facing shared use infrastructure is the design and
performance of critical components. To better understand loads applied to the infrastructure,
UIUC has acquired Wheel Impact Load Detector (WILD) data from sites throughout the United
States from both Amtrak’s Northeast Corridor, (a shared use corridor in operation for many
decades), and the Union Pacific Railroad.

WILD sites are typically constructed on well-maintained tangent track with concrete sleepers.
Although loads experienced elsewhere on the network will vary and may have a higher
magnitude due to track geometry deviations, these data still provide insight to the varied loading
landscape at representative sites throughout North America. Specific loading properties such as
peak vertical load, peak lateral load, impact factor, and speed are analyzed by creating various
distributions of these properties and determining relationships between them. An example of this
type of distribution is shown in Figure 1.

Figure 1: Cumulative Frequency of Peak Vertical Loads on Amtrak at Edgewood, Maryland

(WILD Data from November 2010)

As Figure 1 shows, at Amtrak’s Edgewood, MD WILD site, locomotives, freight wagons, and
passenger coaches all impart different magnitudes of vertical load onto the track structure. Once
the loading spectrum is adequately determined, one must decide how to effectively design the
system and its components accordingly. The relationship between extreme loading events (e.g.
wheel impact loads) and failure mechanisms is not well-defined, so it is difficult to sufficiently
determine the required robustness of design. Probabilistic considerations must be made
throughout the design process, reflecting safety, financial, and capacity decisions. The disparity
in the magnitude of loads between passenger and freight traffic and their respective weighted
traffic volumes must also be addressed in designing for specific loading environments.

Results from the 2012 UIUC International Concrete Sleeper and Fastening System Survey
provide a comparison of the North American and international loading environments and are
summarized in Table 1 [2]. According to both the international and North American responses,
the average maximum freight static axle load exceeds the design axle load based on responses
from the concrete sleeper manufacturers. The load and tonnage values are, on average,
substantially higher in North America than in the remainder of the world, according to the
respondents (Table 1) [2].

Table 1: Loading Environment Summary from the

2012 International Concrete Sleeper and Fastening System Survey [2]

International Responses North American Responses

Average maximum freight 26.8 tonnes (29.5 tons) 35.4 tonnes (39.1 tons)
axle load*
Average maximum passenger 19.6 tonnes (21.6 tons) 26.4 tonnes (29.1 tons)
†
axle load*
Average concrete sleeper design 25.0 tonnes (27.6 tons) 33.9 tonnes (37.4 tons)
axle load
Average annual tonnage 35.1 million gross tonnes 90.8 million gross tonnes
(per track) (38.7 million gross tons) (100.0 million gross tons)
*Interpreted from responses due to discrepancies in axle or wheel loads
†
Light rail response excluded

Both the WILD data and survey results provide a better understanding of the loads imparted into
the superstructure, but this understanding is not sufficient for the design of concrete sleepers and
elastic fastening systems. The load’s attenuation and progression through the track provides
information critical to the design of the superstructure components.

3.2. Qualitative Establishment of Load Path

At their core, mechanistic design practices use actual loading data to develop a design that
functions adequately under the expected loading conditions. To better determine the demands
on each component, an analysis of the load path was conducted at UIUC. An example of this
effort for one specific fastening system and loading scenario is shown in Figure 2. This static
analysis of interface loads and component deflections helped to establish the locations for load
transfer that may require additional analysis.

Given a particular input loading condition and appropriate simplifying assumptions, the magnitude
of forces at each interface can be determined. UIUC has developed a spreadsheet that accepts
particular input parameters, such as material and geometrical component properties, and
produces forces at interfaces and component deflections. Therefore, the spectra of loads such
as those shown in Figure 1 can be traced throughout the remainder of the fastening system (and
the sleeper, ballast, and subgrade), providing estimates of the magnitudes of forces that should to
be measured at each interface given a particular traffic type.
 Figure 2: Concrete Sleeper Fastening System Load Path Map and
Component Free Body Diagram

In addition to this initial analysis, the effect of accelerating wheel loads and clamping force on
longitudinal forces must also be considered in a comprehensive exploration. Because many
simplifying assumptions were used to complete this initial investigation, its results must be used
as an adequate estimation, providing feasible values to be compared with other load
quantification efforts. To evaluate the loads within the system more accurately, lab and field
instrumentation and more sophisticated analyses, such as finite element analysis techniques,
must be employed.

3.3. Laboratory Experimentation, Field Instrumentation, and Analytical Modeling

After identifying locations where the load is transferred throughout the system, it is necessary to
try to accurately quantify the loads that were qualitatively derived. This quantification process
defines the demands on each component, focusing primarily on determining the magnitude of
forces that are transferred at component interfaces. Laboratory experimentation, field
instrumentation, and analytical modeling are tools used to quantify the loading conditions and
displacements at each interface between components.

Both laboratory and field instrumentation provide quantitative information regarding the load path
through the concrete sleeper and fastening system. Using known input loads from full-scale
testing in the laboratory and revenue service testing in the field, UIUC has developed a method
for determining loads applied to the components within the system and their associated
deflections. To correlate the interfacial loads with wheel loads applied at the wheel/rail interface,
significant instrumentation is used on the rail as well. In all, the following magnitudes are
measured in the laboratory and field settings: vertical wheel load, lateral wheel load, vertical rail
strain, rail base bending stress, vertical rail displacement, lateral rail displacement, global vertical
displacement, internal sleeper strain, external sleeper strain, vertical rail seat load, rail seat stress
distribution, insulator post stresses, and fastening clip stress. These values provide a
significantly improved understanding of the behavior of the concrete sleeper and fastening
system as a whole.

In addition to the instrumentation performed on the physical system, significant three-dimensional

(3D) analytical methods are also employed. Using the qualitative free body diagrams as shown in
Figure 2 as a framework and basic statics principles, a fundamental analysis is performed to
determine estimated loads and deflections of the components. Simplified two-dimensional (2D)
finite element models are created to confirm the basic analysis and provide further guidance to
the forces present within the system. In parallel with both the instrumentation and basic analysis,
a comprehensive finite element model is created incorporating the geometry and materials of
each component and its interaction with those surrounding it. This tool can model different
loading scenarios, including dynamic loads, and provide valuable insight into the component
response and interdependencies. Parametric analyses are performed, guiding our understanding
of component properties and how they relate to the performance within the expected loading
regime. Once validated, the model will ultimately be the primary tool for running iterations that
will facilitate the development of mechanistic design practices.

3.4. Design Process

After gaining an improved understanding of the loading environment, one must look at the current
geometry and material properties of the components and evaluate whether or not those
properties are appropriate for the existing and expected loading environment. If not, alternative
component geometries or materials that perform better in response to the loading demands
should be pursued.

The next step in the design process is to relate the loading conditions to specific failure modes.
This is done by identifying certain types of failure that occur specifically because of the loading
demands on that particular component. Taking advantage of the modeling techniques, innovative
designs can be developed and tested using the instrumentation plan already in place. Some
novel component designs are evaluated and existing geometry and materials can continually be
improved. Ultimately, this process will lead to improved mechanistic design practices. This set of
recommendations will be based on both theoretical and empirical relationships, leading to a more
thorough understanding of the behavior and performance of each component.

4. Conclusions

The complex loading conditions found within the concrete sleeper and fastening system in North
America were not fully considered when the AREMA recommended practices were developed.
An improved understanding of the existing loading environment will provide greater insight into
failure mechanisms. The cause of these failure modes can be addressed by improvements to
design recommendations based on the science of those mechanisms. Ultimately, the
mechanistic process of design will lead to improved performance of concrete sleepers and
fastening systems, increased safety, and decreased life cycle costs.

5. Future Research

As this research continues to progress, further insight will be gained regarding the complex
loading conditions present in the concrete sleeper and fastening system. Additional fastening
systems will be tested and analyzed to better reflect their use throughout the railway industry.
WILD data from North American railways will continue to be analyzed and relationships will be
developed, creating a more complete picture of loads being applied to the infrastructure. This
information will be applied to instrumentation and modeling efforts currently underway at UIUC,
continually improving the understanding of the existing environment causing critical failures.
Improved recommended design practices will contribute to the mitigation of these failure modes
and should improve performance of concrete sleepers and fastening systems in North America.

6. Acknowledgements

Funding for this research has been provided by the United States Department of Transportation
(US DOT) Federal Railroad Administration (FRA). The published material in this report
represents the position of the authors and not necessarily that of DOT. Industry partnership and
support has been provided by Union Pacific Railroad; BNSF Railway; National Railway
Passenger Corporation (Amtrak); Amsted RPS / Amsted Rail, Inc.; GIC Ingeniería y
Construcción; Hanson Professional Services, Inc.; and CXT Concrete Ties, Inc., an LB Foster
Company. For providing direction, advice, and resources, the authors would like to thank Conrad
Ruppert, Jr. and Mike Tomas from Amtrak; William GeMeiner from Union Pacific Railroad; and
Winfred Boesterling from Vossloh. Additionally, the authors thank the members of AREMA
Committee 30, Subcommittee 4 (Concrete Crosstie Technology) for their continued support and
guidance in UIUC’s concrete sleeper research. The authors’ gratitude is also expressed to
Andrew Stirk from UIUC, who has provided invaluable service in data processing. The authors
would also like to thank Bassem Andrawes, Zhe Chen, Justin Grassé, Ryan Kernes, Daniel
Kuchma, David Lange, Christopher Rapp, Moochul Shin, Amogh Shurpali, and Sihang Wei from
UIUC for their involvement in this research effort. J. Riley Edwards has been supported in part by
grants to the UIUC RailTEC from CN, CSX, Hanson Professional Services, Norfolk Southern, and
the George Krambles Transportation Scholarship Fund.

7. References

[1] American Railway Engineering and Maintenance-of-Way Association, AREMA Manual for
Railway Engineering, Landover, Maryland, 2012

[2] B.J. Van Dyk, M.S. Dersch, J.R. Edwards, International Concrete Crosstie and Fastening
System Survey – Final Results, University of Illinois at Urbana-Champaign, 2012

[3] European Committee for Standardization, European Standard, Brussels, Belgium, 2009

[4] Standards Australia International, Australian Standard, Sydney, Australia, 2003

[5] W.W. Hay, Railroad Engineering, John Wiley & Sons, Inc., New York, 1982

[6] Japanese Standards Association, Japanese Industrial Standard, Tokyo, Japan, 1997

[7] ARA, Inc., ERES Consultants Division, Guide for Mechanistic-Empirical Design of New and
Rehabilitated Pavement Structures, Champaign, Illinois 2004

[8] J.C. Zeman, Hydraulic Mechanisms of Concrete-Tie Rail Seat Deterioration, University of
Illinois at Urbana-Champaign, 2010