Conference PaperPDF Available

Experimental investigation into the failure behaviour of insulated rail joints

Authors:

Abstract and Figures

Ratchetting failure of railhead material adjacent to the endpost which is placed in the air gap between the two rail ends at insulated rail joints causes significant economic problems to the railway operators who rely on the proper functioning of these joints for train control using the signalling track circuit. The ratchetting failure is a localised problem and is very difficult to predict even when complex analytical methods are employed. This paper presents a novel experiment technique that enables measurement of the progressive ratchetting. A special purpose test rig was developed for this purpose and commissioned by the Centre for Railway Engineering at Central Queensland University. The rig also provides the capability of testing of the wheel/rail rolling contact conditions. The results provide confidence that accurate measurement of the localised failure of railhead material can be achieved using the test rig.
Content may be subject to copyright.
Conference On Railway Engineering
Brisbane 10 12 September 2012
EXPERIMENTAL INVESTIGATION INTO THE FAILURE BEHAVIOUR OF
INSULATED RAIL JOINTS
Paul Boyd 1,a,3, Nirmal Mandal 1,b ,Thaminda Bandula 2,a,3, Nannan Zong2,a,3 and
Manicka Dhanasekar 2,b,3
1 Central Queensland University, Centre for Railway Engineering,; 2 School of Civil Engineering and
Built Environment, Queensland University of Technology; 3CRC for Rail Innovation
a BEng; b PhD
SUMMARY
Ratchetting failure of railhead material adjacent to the endpost which is placed in the air gap between the
two rail ends at insulated rail joints causes significant economic problems to the railway operators who rely
on the proper functioning of these joints for train control using the signalling track circuitry. The ratchetting
failure is a localised problem and is very difficult to predict even when complex analytical methods are
employed. This paper presents a novel experimental technique that enables measurement of the
progressive ratchetting. A special purpose test rig was developed for this purpose and commissioned by the
Centre for Railway Engineering at Central Queensland University. The rig also provides the capability of
testing of the wheel/rail rolling contact conditions. The results provide confidence that accurate
measurement of the localised failure of railhead material can be achieved using the test rig.
1. INTRODUCTION
The purpose of insulated rail joints (IRJs) is to
electrically isolate track sections, allowing the
railway signalling system to locate trains by using
a rail short circuiting method. IRJs are subjected to
variable amplitude dynamic loads generated by
wheel-rail interaction and degrade faster than
other track components [1]. IRJ degradation incurs
significant costs for the rail industry as the affected
joints exhibit early failure due to railhead
ratchetting which affects the structural integrity of
the track system [2, 3].
There are many failure modes of IRJs apart from
ratchetting of railhead material; bolt-hole failure,
mechanical fatigue of the joint bar, and crushing
and delamination of the IRJ endpost are a few.
Degraded IRJs diminish the reliability of the
signalling system; if railhead metal flow causes an
electrical connection between track circuit
sections, this compromises the signalling system
and can cause a serious disruption to train
operations. Therefore, there is a pressing need to
closely examine the failure mechanisms of these
types of joints with a view to improving their
service life, reliability and efficiency. Once the
failure modes are clearly understood, the effects of
design and operational parameters on IRJs can be
assessed to enhance their service life.
Only limited literature is available that addresses
stress and pressure distribution at IRJs using finite
element (FE) simulation, and there is limited
quantitative information on the failure of the
railhead in the vicinity of IRJ rail ends/endposts
using laboratory or field testing [4]. The focus of
this study is the high stresses at the rail ends
producing metal flow shown in Figure 1 that
eventually closes over the top of the endpost and
provides an electrical path, causing failure of the
track signalling system.
Figure 1: Railhead metal flow at the endpost
Mandal [5] performed experimental work on the
study of the metal flow of the railhead adjacent to
the end post using wheel loads causing only
vertical impacts. This work has provided some
useful correlation to support his FEA analysis
results which fail to fully produce the effect of the
rolling contact conditions causing the metal flow.
In this paper, the experimental testing was
performed by replicating wheel-rail rolling contact
Boyd, Mandal, Bandula, Dhanasekar Experimental investigation into the failure
CQU & QUT, Australia behaviour of insulated rail joints
Conference On Railway Engineering
Brisbane 10 12 September 2012
interaction in the laboratory; a test rig for this
purpose was specifically designed. This testing
applies repeated rolling vertical wheel loads. The
vertical unsupported face of the rail in the endpost
gap of the IRJ is the source of the high ratchetting
strains; as this section is difficult to instrument due
to the narrow gap, it was decided to test half of the
full IRJ as shown in Figure 2. This testing was
performed over sufficient cycles to develop
railhead metal flow at the free rail end. The
significance of this work is that the results of this
experimental testing allow the validation of rolling
contact computer models for a rail end to provide a
better understanding of this mechanism of IRJ
failure.
It is recognised that this testing will not develop
impact loading due to the nature of the method
used. However, it has been shown that the effect
on the vertical unsupported railhead face is more
pronounced than the impact loads generated at
the gap of the IRJs. For example, a 130kN wheel
load stationary at the gap produced an equivalent
strain of 17000 microstrain, which is 26% larger
than the same load stationary at a continuous rail
section away from the IRJ gap (where the
maximum equivalent 13500 microstrain occurred).
The same wheel when rolling at 80km/h produced
just 2% larger strain in continuous rail or 19%
larger strain at the vertical unsupported railhead
face), both being less than the dramatic 26%
increase cited above. Considering these results, it
appears sensible to analyse the effect of the
unsupported rail face on the accumulation of
strains as a first stage of examination.
Figure 2: Wheel at free end of a half IRJ
2. WHEEL- RAIL ROLLING CONTACT TEST RIG
The test rig provides a general purpose wheel-rail
rolling contact testing capability with a maximum
vertical wheel rolling load of 300kN and a
horizontal rolling distance of 200 mm. Figure 3
shows the general arrangement of rig
components. The 850 mm diameter wheel (Item 5)
has a cylindrical rolling surface and freely rotates
on plain bearings supported in yokes (Items 4 & 6)
connected to two 500kN servo hydraulic actuators
(Items 2 & 8) which provide the vertical loading
and horizontal wheel movement. The vertical and
horizontal actuators are connected to a steel frame
by support clevises (Items 1 &9). The vertical load
is measured by a fatigue rated load cell (Item 3)
located on the rod of the vertical servo actuator
and the horizontal position of the wheel is
measured with a displacement transducer integral
to the horizontal servo actuator. The horizontal
load is also measured by a load cell (Item 7). The
lateral alignment of the wheel is constrained by
four sliding bearings. Two bearings are located at
the axle ends and a further two are located above
these. Test rails (Item 10) may be supported on
either continuous or discrete supports and at
variable sleeper spacing (Item 11). Rail support
parameters such as angle of tilt, to provide
transverse rail loading and varied L/V ratios can
also be provided. A test rig photograph is provided
in Figure 4.
A digital control system commands the vertical
load values and wheel position as required for
testing programs, and rail loading trajectories may
be either theoretical or those obtained from
captured in-service loading conditions. The testing
cycle rates achievable with the rig depend
primarily on the horizontal wheel movement
requirements and inertial considerations. For
short horizontal strokes, cycle rates of the order of
1 cycle per second have been achieved. For the
purposes of this testing program, the loading cycle
simulated the unidirectional loaded wheel
movement experienced by most IRJs in heavy
haul service where loaded trains travel in one
direction only. To achieve this, the 130kN service
wheel load was applied throughout the forward
movement of the wheel towards and past the free
rail end. When the wheel reached the end of its
travel at a point 10mm past the vertical
unsupported face of the rail (and with the wheel
still in contact with the free end railhead top edge),
the wheel load was then reduced to a low level
load of 5kN before commencing the backward
wheel movement. The test rig was funded through
the CRC for Rail Innovation’s Project R3.100,
“Longer Life Insulated Rail Joints” and
commissioned at the Centre for Railway
Engineering, Central Queensland University.
Figure 3: Test rig general arrangement
Boyd, Mandal, Bandula, Dhanasekar Experimental investigation into the failure
CQU & QUT, Australia behaviour of insulated rail joints
Conference On Railway Engineering
Brisbane 10 12 September 2012
Figure 4: Test rig photograph
3. IRJ RAILHEAD METAL FLOW STUDIES
Finite element analysis (FEA) of the IRJ rolling
contact condition was conducted by Queensland
University of Technology to determine the
complicated stress condition at the rail free end of
the railhead using an Elastic/Plastic model. Wheel
loads were modelled for the wheel rolling along the
rail towards and continuing 10 mm past the free
end. As explained earlier, only half of the IRJ was
modelled as shown in Figure 5 with a view to
numerically simulating the proposed experiments.
Figure 5: Rolling contact FEA model
Modelling the wheel movement past the free end
was considered as a means of achieving
accelerated ratchetting of the railhead. Wheel
positioning past the unsupported edge reduced the
contact area, increased the peak contact pressure
and contributed to higher rates of ratchetting
strains at the top of the railhead. The accelerated
development of railhead plastic flow provided cost
savings for the testing program by reducing the
number of load cycles required to achieve the
desired metal flow. Figure 6 shows the sign
convention used for indicating the wheel position
(L) about the free rail end. The rail end vertical
plane has a coordinate of L=0 mm, indicating that
the wheel axle centreline position is vertically
above the rail end. Positive wheel positions
indicate axle locations over the rail and negative
positions indicate the axle position being located
past the rail end.
Figure 7 provides plots of the values of vertical rail
strain (shown on the y axis) for positions on the
centreline of the free rail end below the rail top
(shown on the x axis). Each curve represents a
different axle position (i.e., different values of L).
Figure 6: Positive and negative wheel positions
Figure 7: Rail free end vertical strains
These strains are presented as they are
considered the only option for measurement of rail
end material strains during test rig experiments for
comparison and validation of the FEA analysis
results. This is due to the inherent difficulty in
directly measuring the rail contact strains.
Test
specimen
Digital
camera
Horizontal
actuator
Steel portal
frame
Vertical
actuator
Wheel
Boyd, Mandal, Bandula, Dhanasekar Experimental investigation into the failure
CQU & QUT, Australia behaviour of insulated rail joints
Conference On Railway Engineering
Brisbane 10 12 September 2012
4. EXPERIMENTAL PROCEDURE
4.1 Strain Measurement
Due to the concentrated nature of the highly
stressed region close to the railhead, it is difficult
to measure strains using strain gauges at any
number of positions in this region, even when
using the smallest strain gauges available. To
overcome this problem, a digital image correlation
technique (DIC) was used in the experimental
studies. A special purpose package developed by
White et al. for geotechnical materials, known as
particle image velocimetry [6], was applied to
determine the strain field throughout this highly
stressed region using successive digital images.
Bandula-Heva and Dhanasekar [7] used this
approach to examine the stress-strain
characteristics of railhead steel. The localised high
strain, wheel-rail interface area is difficult for strain
gauging due to metal debonding problems faced
earlier [5]. A sequence of digital images was taken
of the rail end during the testing cycle and these
images were later processed to derive the strains
at points in the region of interest. In order to
confirm the accuracy of this technique, strain
gauges were however located at the positions
shown in Figure 8. The topmost strain gauge
centre was at 3mm below the top surface of the
railhead, although the analysis shows (Figure 7)
the top 2mm zone is critical. The gauges were
numbered from 1 at the rail top to 8 at the rail
bottom.
Figure 8: Strain gauge locations on rail end
face
Photographs required for the DIC method were
taken with a Canon EOS7D Digital SLR camera
fitted with an Infinity K2/SC Long Distance Video
Microscope lens capable of producing 30 mm wide
images at a working distance of 800 mm, thus
allowing high resolution photographs of the
railhead to be obtained. Figure 9 shows the
camera located at the rail end, and Figure 10
shows one of the images. The camera was
positioned on the rail axis, was focused on the rail
end and was flanked by 2 x 1000 watt narrow
beam lights to provide sufficient image exposure.
Figure 9: Camera setup
The timing of the photographs was handled by the
same control system driving the test rig actuators;
this controlled camera triggering and the creation
of a database providing the associated test
parameters for each photograph.
Figure 10: An image used for DIC analysis
4.2 Loading Trajectories
The specimen loading trajectories consisted of
running 100 cycles at a range of load levels, and
incrementally increasing loads up to the maximum
rig capacity of 300kN. Each load cycle consisted of
firstly positioning the horizontal actuator at the
start point position (L = +20 mm), then increasing
the vertical load to the required high level, then
advancing the horizontal actuator to the required
end position (L= -10mm). The load was then
reduced to the low level (5kN) and the horizontal
actuator retracted to the start point while
maintaining the low load. See Figure 11 for one of
Illumination
Camera
supporting frame
Boyd, Mandal, Bandula, Dhanasekar Experimental investigation into the failure
CQU & QUT, Australia behaviour of insulated rail joints
Conference On Railway Engineering
Brisbane 10 12 September 2012
the vertical load and horizontal wheel position
trajectories used during testing.
Figure 11: Load and position trajectories
One of the test series involved increasing the load
from 50kN to 300kN as shown in Figure 12,
including the 600 loading cycles for the entire test.
Figure 12: Test of vertical loading cycles
A total of four IRJ specimens were tested,
consisting of two rails with joint bars and two
without joint bars to confirm the effect of joint bars
on the results. Varied loading conditions were also
provided Space prevents the presentation of the
additional test outcomes in this paper.
5. EXPERIMENTAL RESULTS
The experimental data and trends are shown in
this section.
5.1 Vertical Strain Variation Along Depth of
Unsupported Rail End
Figure 13 shows the vertical strains along the full
depth of the unsupported rail end. Eight strain
gauges mounted on the rail end face were used to
plot the variation of strain. A 130kN wheel load
located 10 mm past the rail end (at the position of
L = -10mm in Figure 6) is used to plot the strain
profile. It can be seen that the high strain values
occur near the railhead contact region, and there
is a rapid reduction in strains with increasing
depth.
Figure 13: Vertical strains for 130kN Load at
load position L= -10mm
5.2 Validation of DIC
The application of DIC techniques for strain
measurement at the rail end surface is new.
Therefore, it is necessary to validate the DIC strain
results using the strain gauge data as well as the
FE model. More importantly, because strain
gauges cannot be positioned too close to the
railhead top surface due to the high possibility of
delamination, it is advantageous to use the DIC
technique which is able to analyse the strain
components at various locations on the rail end
face. As the DIC method can determine all three
strain components at the unsupported rail end at
any selected point, it offers significant
improvement over strain gauges that usually
measure the strain at a single point in only one
direction.
To validate the feasibility of using DIC for strain
measurement at the rail end, the comparison of
strains amongst the FE results, strain gauge
measurements and DIC derived data are
presented. Figures 14 and 15 present the vertical
strains and shear strains respectively under the
130kN wheel load as the wheel is positioned over
the rail end (L=0). There is good agreement
among the FE, strain gauge and DIC results.
It can seen from Figure 14 that, as the wheel
approaches the rail end, both FE and DIC show
that the maximum vertical strain is shifted toward
the top surface of the rail end. Both FE and DIC
indicate that the strain at a depth of 1mm below
the railhead top surface is valued around 17300,
almost double that at 3mm from the top surface
where the topmost strain gauge could safely be
located. The significant difference in stress in the
top few millimetres of the railhead suggests that
20
15
10
5
0
-5
-10
Boyd, Mandal, Bandula, Dhanasekar Experimental investigation into the failure
CQU & QUT, Australia behaviour of insulated rail joints
Conference On Railway Engineering
Brisbane 10 12 September 2012
the DIC technique is essential to perform the strain
analysis of that region which strain gauge
measurement methods cannot access.
Figure 15 presents the shear strain distribution
along the axis about 7.6mm eccentric to the rail
end symmetric axis. DIC results at the selected
positions along this axis have good agreement
with the FE model. It accurately predicts that the
maximum shear strain is located at around 3mm
below the top surface, valued at around 6200.
Based on the comparison and discussion above,
the DIC method is well validated. Hence it is
considered reliable to use for conducting analysis
of the material plastic strains accumulated under
cyclic wheel loading.
(a) FE prediction of vertical strain for 130kN
wheel load positioned at L= 0mm
(b) Comparison of FE, DIC and strain gauge
data for vertical strain
Figure 14: Vertical strain profile
(a) FE prediction of shear strain for 130kN
wheel load positioned at L= 0mm
(b) Comparison of FE and DIC for shear
strain
Figure 15: Shear strain profile
5.3 Accumulation of Plastic Strain
Figure 16 shows the progressive accumulation of
plastic strain for strain gauge 2 during each load
step. It can be seen that the rate of accumulation
was highest during the initial period of the wheel
passage load increases (from 50kN to 150kN)
than the next load steps.
Symmetric axis
of rail end
Rail he ad
contact surf ace
0
2
4
6
8
10
12
14
16
18
20
-0.02-0.015-0.01-0.0050
Dept h along sy mmet ric axi s
Vertical strain
130KN FE
130KN DIC
130KN Str ain gauges
Symmetric
plane of rail
end
Axis along the
maximum
shear strain
0
2
4
6
8
10
12
14
16
18
20
-0.01-0.008-0.006-0.004-0.0020
Dept h along sy mmet ric axi s
Shear st rain
130KN FE
130KN DIC
Boyd, Mandal, Bandula, Dhanasekar Experimental investigation into the failure
CQU & QUT, Australia behaviour of insulated rail joints
Conference On Railway Engineering
Brisbane 10 12 September 2012
Figure 16: Strains for gauge 2 during
testing
The plastic strain development for a small number
of load cycles is shown in Figure 17. It indicates
that rail end strain begins to increase significantly
for a wheel position of approximately 10 mm
beyond the free rail end. The incremental increase
in plastic compressive strain for each cycle is
clearly evident.
Figure 17: Strain development with wheel
positions for gauge 2
6. CONCLUSION
The rolling contact test rig has proved to be an
effective research tool in providing cyclical rolling
contact loading of the rail and supporting the study
of metal flow in insulated rail joints. Additional
experimental work is being performed on the study
of other aspects of IRJ failure modes and the
testing of complete IRJs. The successful trial use
and strain gauge validation of the DIC method for
non-contact strain measurement has allowed
researchers to overcome the limitations of
traditional strain gauging techniques which rely on
the position of the strain gauges being in the
correct locations to measure peak strains. It is
now considered that the DIC methods ability to
provide strain measurements at any position of
interest on the surfaces of items under test will
greatly assist research work.
In particular the unsupported railhead edge within
the gap of the IRJ was found to accumulate
significant plastic strain during the passage of
each wheel, with the rate of accumulation being
the highest during the initial period of the wheel
passage. The top corner of the railhead edge is
the most vulnerable. The DIC method can
accurately determine the critical strain
components.
ACKNOWLEDGEMENTS
The authors are grateful to the CRC for Rail
Innovation (established and supported under the
Australian Government's Cooperative Research
Centres program) for the funding of this research,
Project No. R3.100, “Longer Life Insulated Rail
Joints”. The authors also acknowledge the support
of the Centre for Railway Engineering, Central
Queensland University, the Queensland University
of Technology and the ARTC and QR industry
partners that have contributed to this project.
REFERENCES
[1] Davis DD, Collard D, Guillen DG. (2004).
Bonded insulated joint performance in mainline
track. Technology Digest. Vol. TD04-006.
[2] Davis DD, Akthar MN. (2005). Improving the
performance of bonded insulating joints. Railway
Track and Structures. January: 14 - 17.
[3] Wen Z, Jin X, Zhang W. (2005). Contact-impact
stress analysis of rail joint region using the
dynamic finite element method. Wear. 258: 1301-
1309.
[4] Nicoli E, Dillard DA, Dillard JG, Campbell J,
Davis DD, Akthar M. (2011). Using standard
adhesion tests to characterize performance of
material system options for insulated rail joints,
Proceedings of the Institution of Mechanical
Engineers Part F-Journal of Rail and Rapid
Transit. 225: 509-521.
Boyd, Mandal, Bandula, Dhanasekar Experimental investigation into the failure
CQU & QUT, Australia behaviour of insulated rail joints
Conference On Railway Engineering
Brisbane 10 12 September 2012
[5] Mandal NK. (2011). Failure of railhead material
of insulated rail joints. PhD Thesis, Central
Queensland University.
[6] White DJ, Take WA, Bolton MD. (2003). Soil
deformation measurement using particle image
velocimetry (PIV) and photogrammetry.
Geotechnique. 53(7): 619-631.
[7] Bandula-Heva T, Dhanasekar M. (2011).
Determination of Stress-Strain Characteristics of
Railhead Steel using Image Analysis. World
Academy of Science, Engineering and
Technology. 60: 1884-1888.
`.
...  in generally the rail joints' failure is connecting to decreased moment of inertia of the pair of fishplates (compared to applied rail section) [1], and enlarged stress values in the head of the rail that can result plastic deformation, lipping in the rail steel material [2],  the maintenance cost of rail joints is significant high, it is very problematic issue to avoid insulated rail joints from railway tracks (mainly CWR tracks) [1],  it is very serious to select suitable type of glue for glued insulated rail joints [3] [4]. It is important and significant issue specially related to the material and thickness of endpost elements [5] [6] [7] [8]. ...
Article
Full-text available
In this paper the authors detail the possibilities of modelling of finite element method (FEM) of glued insulated rail joints which are applied in railway tracks with continuously welded rails (CWR). A lot of laboratory tests (static and dynamic 3-point bending tests, axial pulling tests) were executed on glued insulated rail joints, the specimens were related to three different rail profiles applied in Hungary: MÁV 48.5; 54E1 (UIC54), 60E1 (UIC60), respectively. The static bending tests with many bay length values were conducted, before and after dynamic (fatigue) tests. 2-D beam models were made in FEM software using semi-rigid hinge as the simplified connection of fishplated glued insulated rail joint. The FEM models were calibrated and then validated with the static vertical displacement values in the middle-bay position measured in laboratory. The model validation was conducted with two methods.
... The results showed that irregularities at the welds were sensitive to the plastic deformation of railhead material. Recently, performance of IRJ assemblies has been investigated in laboratory [28] and field [29] conditions. Research on IRJ joint bars was also performed in laboratory [30] and field [31,32] conditions. ...
Article
Insulated rail joints (IRJs) are safety critical components in the signalling system of railway corridors which provide a break in the continuity of the rail steel to locate trains. IRJs connect the two rail ends at the discontinuity to achieve geometric and mechanical requirements of rail. The bending stiffness of an IRJ is about one third that of continuous rail. As a result, the IRJs, especially those in heavy haul tracks, exhibit early failure predominantly due to ratchetting or alternating plasticity of railhead metal in the vicinity of the endpost insulators. A three-dimensional (3D) finite element numerical simulation is carried out to examine failures of railhead material in the vicinity of the endpost of an insulated rail joint considering high frequency dynamic wheel loading. A dynamic wheel load of 182kN is applied through a contact patch; the distribution of contact pressure is considered using a non-Hertzian formulation. A 12m long global IRJ model and a sub-model for localised analysis are employed. The shakedown theorem is employed in this study. Nonlinear isotropic/kinematic elastic-plastic material modelling is employed in the simulation. A peak pressure load lower than the shakedown limit is considered as the input load. The equivalent plastic strain plot for this load case lower than the shakedown limit demonstrates the railhead damage captured through a localised stress analysis in the vicinity of the endpost using the sub-modelling technique. The sub-surface plastic deformation of railhead material extends down to 8mm from the railhead top surface. The critical crack initiating stress components are at 2mm-4mm sub-surface depth. As such, the railhead material fails due to alternating plasticity through low cycle fatigue. Laboratory tests were performed to verify the simulation results and found that test and simulation results correlated well.
... There has been frequent research to investigate mechanical behaviour and failure of IRJs to increase their life (Kabo et al., 2006, Himebaugh et al., 2008, Sandstrom and Ekberg, 2009, Nicoli et al., 2011, Boyd et al., 2012. However, only a few research studies have been conducted considering inserted rail joints and gapped rail at IRJs. ...
Conference Paper
Full-text available
There are two types of Insulated rail joints (IRJs) in the railway track: inserted and glued. The IRJs in the track are significant components in relation to integrity and safety of rail signalling operations. Because of low values of vertical bending rigidity of IRJs, cyclical dynamic wheel loading forces increase significantly at IRJs, thus accelerating material degradation and damage of their components such as rail ends, endposts, joint bars, etc. As a result, relatively free rail ends are created at IRJs that further amplify dynamic wheel load impacts. Only limited literature are available addressing the free end of rail effects at IRJs targeting stress distributions in the vicinity of the rail joints. To understand clearly the delamination process of inserted endpost IRJs and the damage of the endpost and rail end material, a thorough analysis targeting the modes of failure of both glued and inserted IRJs and subsequent damage of the railhead material is necessary to find ways to improve the joint’s life. In this paper, a 3D finite element analysis (FEA) is carried out to assess damage when the endpost is either initially installed loosely (“inserted” rather than “glued”) in between the ends of the rail at the IRJ or becomes a loose fit due to plastic delamination of a glued joint, or when a gap is created between the rail and endpost due to pull-apart problems as the rails contract longitudinally in winter. Both 5mm and 10mm gaps (inserted endpost IRJs) are considered, using a peak vertical pressure load of 2500MPa applied at one rail end. This peak load ensures a ratchetting failure mode of the rail end material. Plastic deformations and stress distributions in the vicinity of inserted IRJs are quantified using FEA data and compared with that of the glued IRJs (5mm and 10mm endpost thicknesses) to show the free end of rail effects. Residual stress and strain distributions indicate the damage of the railhead material. The progressive damage of the railhead material at the rail ends can be quantified by equivalent plastic strain (PEEQ). The free end of rail effects can be further illustrated by comparing PEEQ for inserted and glued IRJs. The railhead material of 5mm and 10mm inserted endpost IRJs is more sensitive to permanent deformation compared to that of the corresponding glued IRJs. Therefore, inserted IRJs pose an increased potential threat to rail operations in relation to damage of railhead and endpost materials and premature failure of IRJs.
Article
Full-text available
Purpose. The aim was to compare behavior of polymer-composite fishplated and control steel fishplated (type GTI and MTH-P) glued insulated rail joints in railway track. Methodology. After laboratory tests (shear tests of glue materials, 3-point-bending tests, axial pull tests), as well as field inspections, trial polymer-composite and control (steel) fishplated glued insulated rail joints were built into railway tracks with (almost) the same border conditions (rail profiles, cross section parameters, track condition, etc.). The authors summarize in this paper the results of field tests related to polymer-composite, as well as control (steel) fishplated glued insulat-ed rail joints between 2015 and 2018 considering measured data of track geometry recording car and straight-ness tests. Findings. The investigation and diagnostics of experimental (fiber-glass reinforced fishplate) and control (steel fishplate) rail joints (straightness tests, track geometry recording car measurements) are in pro-gress. Originality. The goal of the research is to investigate the application of this new type of glued insulated rail joint where the fishplates are manufactured at high pressure, regulated temperature, glass-fiber reinforced polymer composite plastic material. The usage of this kind of glued insulated rail joints is able to eliminate the electric fishplate circuit and early fatigue deflection and it can ensure the isolation of rails’ ends from each other by aspect of electric conductivity. Practical value. The polymer-composite fishplated glued-insulated rail joints and control steel fishplated rail joints were built into the No. 1 main railway line (Kelenföld-Hegyeshalom) in Hungary at three different railway stations. The accurate time could not be determined when the polymer-composite fishplated glued-insulated rail joints reach the end of their lifetime as the result of previous research. In this article the investigation of deterioration process of glued-insulated rail joints is demonstrated.
Article
Full-text available
A deformation measurement system based on particle image velocimetry (PIV) and close-range photogrammetry has been developed for use in geotechnical testing. In this paper, the theory underlying this system is described, and the performance is validated. Digital photography is used to capture images of planar soil deformation. Using PIV, the movement of a fine mesh of soil patches is measured to a high precision. Since PIV operates on the image texture, intrusive target markers need not be installed in the observed soil. The resulting displacement vectors are converted from image space to object space using a photogrammetric transformation. A series of validation experiments are reported. These demonstrate that the precision, accuracy and resolution of the system are an order of magnitude higher than previous image-based deformation methods, and are comparable to local instrumentation used in element testing. This performance is achieved concurrent with an order of magnitude increase in the number of measurement points that can be fitted in an image. The performance of the system is illustrated with two example applications.
Article
Insulated joints (IJs) are often required every few kilometres along railway tracks for signal blocks and rail break detection; practical experience has shown that their life is often a fraction of the life of other track elements on some rail lines subjected to high tonnage freight. This article reports findings from a project conducted to study different bond systems consisting of various combinations of adhesives, fibrous insulators, and rail surface treatments that were of potential interest for increasing the service life of IJs for rail applications. The study was performed in parallel with a finite-element analysis and did not focus on testing real IJs but rather on common adhesion test specimens such as the single lap joint and double cantilever beam configurations. The aim of using these specimens was to simulate potential load and environmental conditions on standard test specimens that were less expensive and easier to construct, test, and analyse. The main goal of the project was to compare a number of combinations of potential IJ components through an extensive test programme. The results highlighted several possible combinations that may warrant further study as actual IJ prototypes. In particular, several material combinations involving materials not currently used by IJ vendors had higher overall performances when compared to currently used combinations, although the extension of improved test specimen performance to actual IJ configurations and service conditions may not follow.
Article
The finite element code ANSYS/LS-DYNA is used to simulate the wheel/rail contact-impact behavior at rail joint region. The implicit and explicit finite element methods have been couple to analyze wheel/rail impact process. In the FE simulation, a material model with linear kinematic hardening was used. The influences of axle load and train speed on contact forces, the stresses and strains in the railhead are investigated in detail. From obtained results, it is found that the axle load has more great effect on the above parameters during impacting at the constant speed. However, the effect of train speed is relatively weak. The study provides a method and useful datum for the further research on fatigue and wear of railhead and improving the rail joint mode.
Failure of railhead material of insulated rail joints
  • Nk Mandal
Mandal NK. (2011). Failure of railhead material of insulated rail joints. PhD Thesis, Central Queensland University.
Improving the performance of bonded insulating joints. Railway Track and Structures
  • D D Davis
  • M N Akthar
Davis DD, Akthar MN. (2005). Improving the performance of bonded insulating joints. Railway Track and Structures. January: 14 -17.
Dhanasekar Experimental investigation into the failure CQU & QUT, Australia behaviour of insulated rail joints Conference On Railway Engineering Brisbane 10
  • Boyd
  • Mandal
  • Bandula
Boyd, Mandal, Bandula, Dhanasekar Experimental investigation into the failure CQU & QUT, Australia behaviour of insulated rail joints Conference On Railway Engineering Brisbane 10 – 12 September 2012