Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Improving aluminothermic weld performance on Transnet’s heavy haul

systems
K. Mistry & J. Duvel
Transnet Freight Rail, Johannesburg, South Africa

ABSTRACT: Improved rail integrity on Transnet’s heavy haul systems is critical in order to ensure reliable,
safe and affordable railway operations. Aluminothermic welds are dominant contributors to rail failures and
related derailments. Improvement to aluminothermic weld performance has a significant influence on overall
rail performance and thus the success of the operations and therefore demand ongoing focus. The objective is
therefore to reduce failures in aluminothermic welds by means of a range of management initiatives. A review
of trends in aluminothermic weld failures, evaluation of root causes and dominant failure modes, together with
detailed metallurgical assessments, drive the important areas of improvement in order to accurately address the
relevant problems. This has led to a wide range of actions which include: effective maintenance management
strategies, training, improved quality control, advanced technologies; and efficient condition monitoring. These
measures have been successful in reducing aluminothermic weld failures.

1 INTRODUCTION 2 BACKGROUND
2.1 Rails
Transnet’s Sishen - Saldanha Iron Ore Export Line
(ORELINE) and the Broodsnyersplaas - Richards Both systems are predominantly installed with
Bay Coal Export Line (COALLINE) demand CEN 60E1 rail profiles of head-hardened grade
progressive development in order to ensure ongoing 350LHT. The remaining construction-vintage UIC60
safe and affordable operations. As aluminothermic and S60 grade 1100 (Cr-Mn) rails are now negligible
welds are dominant contributors to rail failures and within each fleet. New rails are imported in 60 m
related derailments, particularly for the ORELINE lengths, flash-butt welded into 240 m strings in
(Duvel & Mistry 2017), improvement to their Kimberley, and welded into track by means of
performance is central to the reliability of the aluminothermic welding.
systems. Furthermore, the systems are considered to
2.2 Aluminothermic weld process
be critical infrastructure for the nation. The operating
and technical characteristics of each system are given The Goldschmidt Thermit SkV-M aluminothermic
the Table 1. weld process is deployed on both systems due to the
A detailed review of aluminothermic weld more generous gap tolerance, longer pre-heat time
performance has therefore been performed herein, and longer shearing time compared to the SkV-F
followed by root cause analysis and targeted process. The SkV-F process was used previously but
strategies to reduce failures and mitigate derailments. proved to be problematic in terms of high failure
rates. The SkV-M process on the other hand is more
Table 1. System characteristics forgiving, particularly in casting of emergency welds
Parameter ORELINE COALLINE during any time of the day – a huge benefit
Route Length 860 km (Single line) 510 km (Double line) considering the prevailing climatic conditions. The
Train Length 342 Wagons 200 Wagons SkV-M weld process is shown diagrammatically in
Annual Gross Tonnage 84 MGt Line 1: 90 MGt
Figure 1 and the primary differences between the two
FY16/17 Line 2: 28 MGt
Total: 118 MGt processes are summarised in Table 2.
Maximum Axle Load 30 tonnes 26 tonnes The Single Use Crucible (SUC) is used and the
Maximum Speed (Loaded) 60 km/h 80 km/h weld portion is Z-120 for head-hardened rail. No
Ruling Gradient (Loaded) 0.40% 0.63% post-cast cooling treatment is deployed.
Minimum Curve Radius 1000 m 550 m
40
Pre-heating is by means of petrol-air mixture at a ORELINE
combined pressure of 25 - 30 kPa. Petrol-air pre-heat 35 COALLINE
is beneficial due to the following:
 Proximity of petrol availability to remote areas; 30

Number of Rail Breaks


 Mobility of heater equipment; 25
 Fixed hose length resulting in consistent pressure;
20
and
 Improved Welder productivity and unit cost. 15
Furthermore, weld failure rates related to petrol-
10
air pre-heat were found to be considerably lower
compared to the oxygen-propane pre-heat previously 5
used. Oxygen-propane was problematic due to issues
0
with hiring/ transport/ storage of cylinders; poor 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17
availability of suppliers in remote areas; and Financial Year To Date

variability in oxygen quality and tendency for Figure 2. Rail break history (including all weld breaks)
Welders to extend hoses – thereby reducing pre-heat
effectiveness.
3.2 Breaks in aluminothermic welds
For both systems, breaks related to aluminothermic
welds account for almost half of all rail breaks (Duvel
& Mistry 2017). Figure 3 demonstrates the trends in
breaks in aluminothermic welds since 2010. There
has been a dramatic improvement in performance for
the ORELINE from 2012 and this is as a result of a
wide range of successful interventions actioned
during the period. The COALLINE shows some
improvement since 2013, but marginal improvement
over the entire period.

20
ORELINE
COALLINE

15
at Aluminothermic Welds
Number of Breaks

10
Figure 1. Diagram of SkV-M weld process with SUC

5
Table 2. Primary differences between weld processes
Process SkV-F SkV-M
Allowable Gap 24 - 26 mm 35 - 40 mm 0
Pre-heat Time: Petrol-Air 7 minutes 8 minutes 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17
Financial Year To Date
Shearing Time from Pour 5 minutes 6 - 7 minutes
Figure 3. Break history: aluminothermic welds

3 RAIL BREAK TRENDS


3.3 Aluminothermic weld break density
3.1 Rail break history
The comparison of the average density of breaks
The annual trend in rail breaks (i.e. total rail breaks, occurring per year in aluminothermic welds,
including all weld breaks) is shown in Figure 2 for normalised by line length, is given in Figure 4.
each system. Whilst the ORELINE shows significant COALLINE Line 1 (which almost exclusively carries
reduction in rail breaks since 2010, despite gross loaded coal traffic) is therefore currently
tonnages having in fact doubled through the period, experiencing a significantly higher density of breaks
this level of dramatic improvement is not observed on in aluminothermic welds than the ORELINE. This is
the COALLINE. The respective trends are driven by due to the fact that, whilst the ORELINE is operating
the trends in breaks in aluminothermic welds at a 30 tonne axle load, COALLINE Line 1 is
demonstrated below. expected to suffer greater rail stresses – resulting in
reduced safety margin with respect to the weld the ORELINE is considered to be overall more
fracture toughness – due to the following: effective, specifically in managing aluminothermic
 Greater lateral loading and Rolling-Contact weld failures. This is very positive in light of the high
Fatigue (RCF) damage due to higher eccentric derailment probability for the ORELINE (Duvel &
loading as a function of the sharper track Mistry 2017).
curvature;
 Greater rail tensile longitudinal stresses due to the 80%
increased traction loading, braking forces and
70%
related rail movement due to steeper gradients;
and 60%

(Aluminothermic Welds)
 Greater propensity for rail internal defect

Break / Defect Ratio


50%
occurrence and growth due to higher rail surface
damage and contact stresses. 40%

0.011 30%

0.010
20%
0.009
Annual Breaks per Rail km
(Aluminothermic Welds)

10%
0.008

0.007 0%
ORELINE COALLINE Line 1 COALLINE Line 2
0.006 System
0.005
Figure 5. Break / defect ratio for aluminothermic welds
0.004

0.003

0.002 3.5 Types of aluminothermic weld breaks


0.001 At Transnet (Transnet 2012), breaks at
0.000
ORELINE COALLINE Line 1 COALLINE Line 2
aluminothermic welds are classified as being either:
System  Straight type breaks “through” the weld, i.e. within
Figure 4. Density of aluminothermic weld breaks the weld collar: Break codes H2 and H6. Refer to
Figure 6 which indicates the this break
classification code according to Transnet’s
COALLINE Line 2 performance is significantly Manual for Track Maintenance (Transnet 2012);
better than that of Line 1 and the ORELINE. This is or
expected, as the annual traffic volume for Line 2 is  Straight type breaks “adjacent” to the weld, i.e.
less than a third of that of Line 1 (refer to Table 1). beyond the edge of the weld collar, in the heat
Furthermore, Line 2 almost exclusively carries empty affected zone: Break codes H3 and H7; or
coal wagons returning to the coal fields for re-  Horizontal split web type breaks: Break code H5.
loading, whilst Line 1 almost exclusively carries
wagons loaded with export coal at an axle load of 26 For both systems, the straight break type is
tonnes. There is naturally increased fatigue life for overwhelmingly dominant. Straight breaks through
aluminothermic welds on Line 2 due to a lower (break codes H2 and H6), represent three quarters of
propensity for internal defect growth due to the break types in aluminothermic welds and are thus the
combination of lower bending stresses and lower focus of analysis. Only one horizontal split web type
contact stresses. break (H5) has been experienced during the period
(ironically on COALLINE Line 2) and this is an
3.4 Assessment of management of rail breaks extremely isolated case.
The management of rails includes routine rail testing
to detect defects prior to ultimate failure (Duvel
2005). The break / defect ratio, an internationally
recognised metric (IHHA 2001), assesses the
effectiveness of rail management strategies, based on
the premise that ultrasonically detected defects are
necessary for the effective control of rail breaks Figure 6. Aluminothermic weld break classification:
(Duvel & Mistry 2017). Straight break through
As demonstrated in Figure 5, the ORELINE shows
a significantly better break / defect ratio for
aluminothermic welds than the COALLINE.
Therefore, the ultrasonic testing regime in place on
4 ROOT CAUSE ANALYSIS According to the metallurgical analysis, the typical
microstructure of shrinkage cavity in the web was
Rail breaks are routinely sent (ideally within 7 days found to be pearlitic. Here, shrinkage porosity is
of occurrence) to Transnet’s Metallurgical present amongst grains of pearlite, as shown in Figure
Laboratory for detailed analysis. This enables 8. Micro-shrinkage porosity is also present, however
evaluations of the root cause and thorough this is typical of any castings (Campbell & Harding
metallurgical assessment of each aluminothermic 1994).
weld failure. The dominant failure modes for straight Examination under a scanning electron
breaks through the weld have therefore been microscope revealed that the initiation point of the
established. These include: fracture surfaces are typical dendritic structures, as
 Shrinkage cavity in the web; shown in Figure 9. These dendrites indicate either
 Stress raiser defect in the foot; and insufficient heat or insufficient feed of molten metal
 Hot tear in the head/ web area. and this therefore proves that insufficient pre-heat is
influential.
4.1 Shrinkage cavity in the web
Shrinkage cavities in the web typically occur at the
lower end of the web. Refer to Figure 7. The defect
forms due to insufficient liquid metal flow into the
cavity that develops as the surrounding areas solidify.
A microstructure change also occurs and the
combined effect is volume shrinkage and subsequent
contraction (Chen et al. 2006). The typical causes are
all Welder related, as follows:
 Movement occurring during the removal of the
mould shoes; or
 Improper preparation of the weld prior to
stripping; or
 Insufficient pre-heat. Figure 8. Porosity microstructure: shrinkage cavity in the web

This failure mode is more dominant on the


COALLINE due to the more extreme temperature
variations of the area. There is thus propensity for rail
longitudinal tensile stresses to develop during casting
if welds are cast during times of falling temperatures,
and this subsequently contributes to the formation of
shrinkage cavities.
Shrinkage cavities in the web are oriented in the
direction of the zero degree ultrasonic probe used for
the web and are therefore not detectable by means of
the routine ultrasonic testing currently deployed.

Figure 9. Dendritic structure: shrinkage cavity in the web

4.2 Stress raiser defect in the foot


Stress raiser defects in the foot commonly occur at
the rail to weld interface or close to the edge of the
weld collar. Typically, the initiation is related to:
 Lack of fusion due to a cold shut. This is primarily
as a result of: the moulds not being centrally
placed over the gap and/ or the moulds being
skewed due to excessive rail height difference and/
or incorrect weld gap. Alternate causes include the
luting sand not being properly compacted under
Figure 7. Shrinkage cavity in the web the foot. Refer to Figure 10; or
 Lack of fusion due to uneven pre-heat; or
 Flashing. This is due to molten metal solidifying
where the two piece moulds are supposed to join
closely. Sand burns can also occur. See Figure 11.
The causes include the opening of the mould under
the foot due to improper fitment and/ or
overtightening and/ or placing the clamps to high.
There is thus a potential flaw in the use of the two
piece mould design which needs to be effectively
counter-acted with Welder training.

The COALLINE is notorious for this failure mode as


a result of problems with weld preparation. This
applies specifically to the tendency for Welders to
rush the preparation of the weld due to operational Figure 12. Fractograph: Lack of fusion defect in the foot
pressures placed on them to open the line to traffic
sooner than is reasonable. Issues with mould fitment 4.3 Hot tear in the head/ web area
therefore proliferate. Hot tears in the head/ web area occur at the rail to
Stress raiser defects in the foot are not always weld interface at the edge of the weld collar, as shown
situated directly under the web, and are transversely in Figure 13. There may be individual defects on
oriented (i.e. in the direction of the zero degree either the gauge or field side, or on both sides of the
ultrasonic probe) and are therefore not detectable by rail. Typically, the defects are as a result of disruption
means of the routine ultrasonic testing currently of weld solidification due to causes described above
deployed. for shrinkage cavity in the web. Furthermore, it is
possible that negative temperature changes and
resultant rail tensile longitudinal stresses are
sufficient to cause the hot tear.
This failure mode is more dominant on the
ORELINE due to the tendency for the Welders on the
system to plate and clamp the weld while it is still in
the plastic state. The restraint on the weld thus
generates tensile stresses during cooling and this
results in the hot tearing.
The zone of the weld in which these hot tear
defects are situated is typically not visible to any of
Figure 10. Stress raiser defect in the foot due to lack of fusion the ultrasonic probes currently in use. The ultrasonic
detectability of these defects is therefore very low.

Figure 11. Stress raiser defect in the foot due to flashing

The initiation point for a lack of fusion defect in the


foot is characterised by a very small dendritic area,
followed by a fatigue crack, followed by an area of
fast brittle failure. Refer to Figure 12. The scanning
electron microscopy typically shows a smooth
fracture surface. (It is difficult to observe the beach Figure 13. Hot tear in the head/ web area
marks in this image due to the beach marks being
confused with the pearlitic microstructure.)
Examination under a scanning electron 5.2 Radiographic defect types
microscope revealed that the initiation point of the For both systems, the predominant defect type is
fracture surfaces are typical dendritic structures wormholing (85% of radiographic defects), followed
similar to those shown above for shrinkage cavity in by inclusions at 10%. This high incidence of
the web. wormholing is commonplace for the Transnet
environment and is a function of the combination of
5 RADIOGRAPHIC TESTING the physical occurrence of this particular defect and
the defect classification standard currently adopted
All aluminothermic welds are radiographed (“X- by Transnet. Other heavy-haul railways typically
rayed”) within 7 days of manufacture for acceptance, experience higher incidences of inclusions and
as part of Transnet’s established weld quality control significantly less wormholing (Mutton 2007). The
regime. Radiographic inspection is proven to be predominant location of defects reported by the
highly effective in identifying weld defects not radiographic examinations is the foot of the rail,
identifiable by ultrasonic testing (Duvel 2005). which is typical for a heavy-haul railway (Offereins
Furthermore, it has been found at Transnet that & Mutton 2001).
poorly manufactured welds likely to fail the Wormholing is essentially a form of elongated
radiographic examination in terms of internal defects porosity due to gasses drawing during weld
invariably have additional and/ or related defects such solidification. This is affected by various
as dipped vertical profile and/ or breakouts of the rail combinations of the following Welder-dependant
crown. These conditions influence impact loads on issues:
the rail and place the weld under severe dynamic  Inadequate pre-heat of the multi-use crucible;
stress and further increase the risk of weld failure.  Moisture in the mould, crucible or luting sand;
Radiographic testing entails films positions on the  Rails/ moulds not aligned;
web and foot of the rail as shown in Figure 14. One  Sand not applied properly to seal the moulds; and
radiographic film is used for the foot which is  Insufficient pre-heat.
subjected to two exposures on each side of the foot,
whilst the third shot is on the web. Inclusions are foreign particles trapped in the weld,
affected by various combinations of the following
Welder-dependant issues:
 Cleanliness of the crucible;
 Cleanliness of the mould;
 Cleanliness of the rail ends; and
 The intrusion of sand/ mould particles into the
weld metal.

Typical radiographic defects are demonstrated in


Figures 15 to 18.

Figure 14. Radiographic film positions

5.1 Radiographic failure rates


An assessment of the failure rates for 2016 shows that
the overall percentage of welds failed (i.e. Class 2 Figure 15. Radiograph indicating wormholing in rail foot
plus Class 3 failures) is 12% for the ORELINE and
9% for the COALLINE. Transnet’s targeted
radiographic failure rate for the heavy haul systems is
less than 5% and this is certainly possible, as proved
by the very good performance of certain Welders,
some reporting failure rates much lower than 5%.
dominant root cause failure modes for weld breaks
are typically located in such zones of the weld.

6 ADDRESSING THE PROBLEMS


6.1 Welder assessment
A certification system is being established whereby
routine formal training and practical assessment of
Welders is a mandatory safety-critical requirement
done every two years.
Figure 16. Radiograph indicating inclusions in rail foot Welders showing high failure rates are
immediately stopped and re-assessed/ re-certified.
Task observations are done and the Welder’s working
conditions and equipment critically assessed.

6.2 Welder training


A one day on-site workshop/ presentation is
conducted by the approved welding equipment/
consumables supplier, ideally every six months, as a
refresher session. Failure rates, types of defects and
related causes are discussed and welders concerns are
raised. Precision and consistency is imperative
Figure 17. Radiograph indicating shrinkage cativy in rail web
(IHHA 2001). Each step of the welding process is re-
enforced and focus is placed on the relevant standards
and specifications, particularly:

Weld preparation:
 Square cutting of rail ends;
 Rail setup and alignment;
 Gap setting; and
 Accounting for rail wear: Rail height difference
shall not exceed 6 mm and the correct step mould
kits shall be used if rail height difference is 3 – 6
Figure 18. Radiograph indicating lack of fusion in rail foot mm. Where rail height difference exceeds 6 mm,
an appropriate closure to be used. This is critical
as it is a common fault.
5.3 Relationship to failure modes
Fitting of moulds:
Converse to the radiographic defect types, the  Cleaning and adjustment;
dominant root cause failure modes for weld breaks  Mould placed centrally over the gap and
described above (i.e. shrinkage cavity in the web, alignment/ fitment;
stress raiser defect in the foot and hot tear in the head/  Sealing of moulds with luting sand, correctly
web area) do not include wormholing and inclusions. compacted at the bottom of the rail foot. This is
This is due to the fact that the failure modes for weld critical as it is a common fault; and
breaks are typically not detected by radiographic  Fitting of mould shoes: Caution to not overtighten
testing as a result of the following limitations of the as this could lead to opening of the mould under
radiography deployed: the foot.
 There are three major blind spots where it is not
possible to detect defects, i.e. the head/ web area, Pre-heating:
the web/ foot area and the centre of the foot.  Burner to be placed centrally over the gap;
Therefore, stress raiser defects in the foot and hot  Gas pressures; and
tears in the head/ web area are naturally missed;  Pre-heating time.
 Furthermore, the radiographic plate is distorted at
the web/ foot area and thus shrinkage cavities in Tapping
the web are very often obscured as these occur
mostly at the lower end of the web; and Finalising:
 Radiography is less effective for material  Mould removal procedure and timing;
thicknesses of greater than 20 mm. Defects for the
 No rail movement during weld solidification; 7 QUALITY CONTROL
 Stripping and shearing; and
7.1 Radiographic testing
 Final grinding.
Compliance with the stipulated radiographic testing
Accounting for environmental/ weather conditions, regime is being enforced:
particularly humidity.  X-Rays to be done within 7 days of manufacture;
 Reports issued within 14 days of manufacture;
 Failed welds to be replaced (Class 3: Cut out with
6.3 Welder recognition an emergency occupation. Class 2: removed
Formal rewarding of Welders is done regularly within 30 days of manufacture); and
according to low radiographic failure rates. An  Welds to be clamped from the date of
"Annual Award for Welding Excellence" is given to manufacture, until cleared. All clamps to be
Welders who prove consistently low failure rates for inspected weekly and tightened/ adjusted if
the various Depots within each system. necessary.

A documentation system is established to:


6.4 Single Use Crucible
 Properly manage all new welds from installation,
Failure rates have been proven to reduce with the to radiography, to reporting, to final acceptance
adoption of the Single Use Crucible (Hefer & (or replacement if failed); and
Radmann 2015), for the Thermit SkV-M welding  Routinely assess Welder performance according
process deployed. This is due to the superior to radiographic test results.
cleanliness and reduced intensity of preparation
compared to the previously used long life (“multi
7.2 Inspection
use”) crucible. There is thus significantly reduced
scope for Welder error or carelessness with the Single Visual inspection and geometric measurement
Use Crucible and it has therefore been used (especially vertical profile) of all new welds is done
exclusively since 2012. The incidence of breaks in within 7 days of manufacture and recorded.
aluminothermic welds has reduced as a result, Acceptance criteria are as follows:
particularly on the ORELINE.  The weld surface shall be free of any cracks, slag,
inclusions, undercuts, porosity or foreign material;
 The head and foot of the rail on either side of the
6.5 Thermal insulator
weld collar shall be vertically aligned to a
Research has shown that the solidification rate tolerance of 3 mm for the standard mould;
influences the formation of shrinkage defects. It is  The foot of the rail on either side of the weld collar
therefore important to ensure that the cooling rate is shall be horizontally aligned within 3 mm;
controlled. This entails adherence to correct pre-  The two piece moulds must be aligned to a
heating times which enable adequate solidification tolerance of 3 mm;
time for the molten metal. The introduction of a  After final grind, a 0.2 mm feeler gauge may not
thermal insulator at the foot of the rail (refer to Figure fit through any gap between the centre of the rail
19) has proven to ensure a slower cooling rate and crown running surface and a 1 metre straight edge
better mechanical properties of the weld metal placed longitudinally across the weld; and
(Simelane 2012).  A 0.2 mm feeler gauge may not fit through any gap
between the gauge face and a 1 metre straight edge
placed longitudinally across the weld.

8 REDUCING BENDING STRESSES


Due the reduced fracture toughness of welds, it is
imperative to reduce bending stresses where possible
(Gehrmann & Radmann 2013). The following have
therefore been implemented:
 Improving track support. All welds are tamped 7
days from manufacture;
 Monitoring of rail surface condition. Plasser
Figure 19. Thermal insulator between the mould and shoe IM2000 track geometry measurement car data is
being used to identify locations of high wheel-rail
impacts and thus dipped welds, rail break-outs/
spalling, etc. (in addition to other faults such as
faulty/ damaged turnout components). This is by
means of the Vertical Force Reading (VFR) of the various interventions, particularly for the
metric. VFR accelerations exceeding 30g are ORELINE. This is a continuous process requiring
prioritised, followed by site inspections and action ongoing evaluation, assessment and adjustment.
plans to eliminate the defects; and
 Monitoring of vehicle condition. A thorough 10 REFERENCES
review of alarm levels/ hierarches and respective
action criteria for the relevant wayside monitoring Campbell, J. & Harding, R. 1994. Solidification Defects in
equipment, specifically the weighbridges and the Castings, Talat Lecture 3207, European Aluminium
Weigh-in-Motion Wheel Impact Monitor (WIM- Association
WIM), is planned. This will include detailed audits Chen, Y., Lawrence, F., Barkan, C., Dantzig, J. 2006. Weld
of equipment availability/ calibration, and audits defect formation in rail thermite welds. Proceedings –
Institution of Mechanical Engineers Part F Journal of Rail
of alarm monitoring at the Centralised Train and Rapid Transit, December 2006
Control Centre (CTC). Duvel, J., Mutton, P. & Alvarez, E. 2005. Rail requirements for
40 tonne axle loads, Proceedings – 8th International Heavy
Haul Conference, Rio de Janeiro, June 2005.
9 CONCLUSION Duvel, J. & Mistry, K. 2017. Improving rail integrity on the
More than half of all rail breaks on Transnet’s heavy Sishen-Saldanha line, Proceedings – 11th International
Heavy Haul Conference, Cape Town, September 2017.
haul systems relate to aluminothermic welds. There Gehrmann, R., Radmann, P. & Keichel, J. 2013. Optimisation
is therefore a significant impact of the improvements of thermit welding, Proceedings – 10th International Heavy
to aluminothermic weld performance on overall rail Haul Conference, New Delhi, February 2013.
performance and system safety. Hefer, F., Radmann, P. & Keichel, J. 2015. Rail joining under
The dominant type of break is straight and there heavy haul conditions, Proceedings – International Heavy
Haul Conference, Perth, June 2015
are various failure modes which all relate to Welder IHHA, International Heavy Haul Association. 2001. Guidelines
performance and are thus avoidable to a large extent. to Best Practices for Heavy Haul Railway Operations:
The issues of technology/ product selection are thus Wheel and Rail Interface Issues. Chapter 5.2: Rail Structural
secondary to rectifying Welder performance and Deterioration
improving weld management. Mutton, P. 2007. Failure modes and non-destructive inspection
Therefore, ongoing training of Welders and requirements for aluminothermic rail welds under high axle
load conditions, Proceedings – 5th International Institute of
improvements to quality control are central and Welding Asian Pacific International Congress and Welding
safety-critical. The ORELINE requires focused Technology Institute of Australia Technology Week, Sydney,
attention due the more challenging operating March 2007
environment and higher probability of derailment due Offereins, G. & Mutton, P. 2001. Recent experiences with the
to breaks at aluminothermic welds (Duvel & Mistry performance of aluminothermic rail welds under high axle
loads, Rail Track Conference, 2001
2017). Performance on the COALLINE can do with Simelane, N. 2012. The innovation of using the insulation insert
some improvement, particularly for Line 1 due to the as means of reducing heat losses during aluminothermic
rail stress conditions. welding in uncontrolled conditions. Bachelors of
There has been an improvement to Technology dissertation, University of Johannesburg.
aluminothermic weld performance since introduction Transnet. 2012. Manual for Track Maintenance (2012)

You might also like