This poster won the "IIM Springer Best Short Presentation Award'' at the National Metallurgists' Day-Annual Technical Meeting, 2017 held at BITS Goa

1. Indian Institute of Metals, NMD-ATM
71st Annual Technical Meeting
November 11-14, 2017
FAILURE ANALYSIS OF A
TONGUE RAIL
Abhilash Agnihotri, Siddharth Dhananjay, Anirudha
CS, Gurudath B, Preitish D’Silva, Subray Hegde
NITK-Surathkal
s
NMD-ATM-2017R-00445
Point
Introduction
A transverse fracture occurred in a tongue rail on the
19th of July 2016 between ZARP and SWV in the
Ratnagiri region of Maharashtra. The fracture took
place in the tongue rail within the fish plate assembly.
The tongue rail designated ZU49E is used for change
of rail track/lane whose gradually changing cross
section was designed in Germany. There are three
types of joints commonly encountered in the track,
namely,
• Bolted joints : Used to join two rails in jointed rail
territory
• Compromise joints: Used to join two rails of
different cross sections
• Insulated joints: Can be bonded (glued) or non-
bonded (a bolted joint with insulating properties)
By their very nature these joints create discontinuities
and could accelerate track degradation. The gap
between tracks is a source of impact loading from
passing wheels. The failed track was part of a turnout.
The constituents of a turnout are shown in the figure
below,
Fracture in rails at critical locations such as joints,
track changes etc., have become prevalent and have
caused derailments in many cases. The Brétigny-sur-
Orge train crash of 2013 which occurred in the
outskirts of Paris, France left 7 dead with around 200
injured. In as recent as 2015, the transportation safety
board of Canada released a report of a derailment
which occurred near Dublin, Ontario. Most of these
failures involved cracks initiating from bolt joints and
so the importance of securing rail joints cannot be
understated.
Switch Rod
Stock Rail
Closure Rail
Frog
Guard
Rail
Point
Figure1: Constituents of a turnout
The objectives of this study are,
1. To examine the failed samples thoroughly and
look for failure signatures.
2. Perform non destructive testing to identify the
origins of cracks.
3. Propose a mechanism of failure based on visual
inspection.
4. Verify the mechanism using Finite Element
Analysis
5. Suggest modifications to avoid failure in the
future.
Objectives
(all in wt%)
Experimental
Failure Analysis:
a. Material and Microstructure:
The rail material is a wear resistant high carbon
steel whose nominal composition is given in the
table below. The microstructure is 100% pearlite as
is evident from the microstructure shown below.
b. Visual Inspection:
c. Fractography
d. Dye Penetrant Inspection
Fig1: Microstructure of the rail with 100% pearlite
Fig2: Side view of the as received tongue samples
Fig3: Bolt shank wear Fig4: Separated head
Fig5: Horizontal fracture surface in the web showing crack initiation site
(arrow), fatigue beach marks and brittle river bed pattern
Fig6: Angular fracture surface showing crack initiation site at the bolt hole
(arrow) and beach marks
Fig7: Close up of the bolt hole before and after dye penetrant inspection
Two other cracks
Through thickness crack
C Mn Si S P Al Mo Fe
0.6-0.8 0.8-1.3 0.1-0.5 0.03
(max)
0.03
(max)
0.015
(max)
0.04
(max)
Balance
Proposed Failure Mechanism
1. Off-centric loading because of the turnout
2. Impact load is also involved
3. Bolt hole is a stress raiser in general.
Further worsened by slack in the bolt
4. Cause of head-web separation=Off-centric
loading
5. Cause of bolt hole cracking=Slack in the
bolt
all in wt%
Results
I. Simulation 1:
Loading and Boundary conditions:
1. Force = 107 kN applied symmetrically over
the track surface.
2. For this model, only the base is considered
as a fixed support.
3. Edge sizing used in the model is 1mm in
the bolt hole. Automatic meshing, which is
the default condition, is used for the rest of
the track.
II. Simulation 2:
Loading and Boundary conditions:
1. The weight component of the load (107
kN) was applied off centrically to the
curved surface of the track. The impact
component of the load (1800 N) was
applied to the edge of the track where
impact occurred.
2. Edge sizing of 1mm was used for the bolt
hole.
3. The base of the track, the face opposite to
that which took the impact and the bolt
were all taken as fixed supports.
Fig8: Simulations showing von Mises stress in the rail and bolt-hole
Fig9: Comparison of the failed specimen with simulation
Conclusions and Suggestions
Visual inspection, DPI and fractographic analysis
confirmed head-web separation and bolt hole
failure to be fatigue failures. Bolt shank wear lead to
the conclusion of the bolt being loose. Off-centric
loading was proposed to be the primary cause of
head-web separation and was proved by the
simulations that were carried out. Chamferring of
bolt holes, Wheel Impact Load Detectors, tightness
check for bolt holes may avoid further recurrences.
References
1. Railway Investigation Report (2007), R07Q0001, Transportation
Safety Board of Canada.
2. Railway Investigation Report (2015), R15H0005, Transportation
Safety Board of Canada.
3. Gurudath B, (2016). Stress Analysis of Rail Steel. M.Tech thesis,
NITK Surathkal.
4. Subray Hegde. (2016) Failure of KRCL Tongue Rail. NITK
Surathkal