2019 Annual Review

Presentations

©2019 TTCI
TTCI is a wholly owned subsidiary
Slideof1the AAR
of 24
Track and
Structures
Evaluations:
FAST and
Revenue Service
Joseph LoPresti
Megan Brice
Ben Bakkum

©2019 TTCI
TTCI is a wholly owned subsidiary
Slideof1the AAR
of 19
Track Component and Structures Evaluations
• Increase understanding of effects of high tonnage and heavy axle loads
 Heavy trains accelerate track degradation
 Well-planned evaluations provide important information
- Product development and improvement
- Industry implementation
 Reduced risk and increased safety
 Less maintenance and downtime increased capacity and efficiency

Slide 2 of 19
Evaluation and Implementation

Eastern Tests
NS Railroad Concept/
Modeling/
Lab Testing

Viable idea
New Design
Meets FAST FRA Owned,
Improvements
Minimum
or AAR Operated
Design
Alternatives
Requirements

Successful
Tesing in
Controlled
Environment
Western Tests Northern Tests
UP Railroad CN Railroad

Slide 3 of 19
Key Characteristics of the Test Sites
FAST Western Eastern Northern
Car weights (lbs) ≈90% 315k ≈80% 286 k ≈60% 286k ≈60% 286k
≈10% 286k ≈20% 263k ≈40% 263k ≈40% 263k
Annual Tonnage 130 – 160 MGT 120 – 140 MGT 40 – 50 MGT ≈110 MGT Site 1
≈40 MGT Site 2
Curves 5 and 6 degrees 1 and 2 degrees 9 to 11 degrees Up to 5 degrees
Train Speed 40 mph 40 – 50 mph 25 – 35 mph 30 – 40 mph
Track Structure • Concrete ties Concrete ties with Hardwood ties Concrete ties with
• Wood ties with mix elastic fasteners with elastic elastic fasteners
of fasteners fasteners
• EPC ties with mix of or cut spikes
fasteners
Site • Semi-arid western Mid-western plains Eastern Varied Canadian
plains mountainous terrain. More than
200 days/year below
freezing
Slide 4 of 19
Facility for Accelerated Service Testing (FAST)

• 2.7-mile High High Tonnage Loop

Tonnage Loop
• Mostly 315,000 Thermite
Welds
Special
Trackwork

pound cars Special

Rainy
Section

• ≈50% of train Trackwork

Ties &

operations in
Fasteners
Special
Rail Trackwork
each direction Rail
Ties & Bridges

• Over balance Fasteners

Bridges
Bridges

speed in curves

Slide 5 of 19
 Rail Tests

• High strength rails

 2014 to 2018: 650 MGT – differences EFB weld break
between suppliers in wear and RCF initiation point
 Eight electric flash butt (EFB) weld breaks
 New HS rail test started in 2018

RCF Cracks on rail head surface

Slide 6 of 19
Rail Tests

• Intermediate strength
rails – effects of wide
and narrow gage
 Rails installed at 56 1/4-in.
and 57-in. gage
 Minor differences in wear
and RCF patterns
 TTCI’s models being used
to understand causes

Slide 7 of 19
Rail Welding

• Thermite welding
 Full section higher-
hardness welds
performing well
through 160 MGT
 Twelve weld collars
treated with
ultrasonic peening.
Tool improved
based on user input

Slide 8 of 19
Bridges

• Five riveted steel spans

 Four 100+ years old
 FAST train exceeds normal
rating on four spans
 One cover plate notched to
promote crack growth
 All spans performing well
• Two concrete spans
 One, box girder, one slab
 Both spans performing well

Slide 9 of 19
Special Trackwork

• Two #20 turnouts, and

one #11 frog
 Geometry for improved
steering, heavy point frog for
Class 6 track, optimized
stiffness (#20)
 Laser clad frog repair
(ended early 2018)
 New design flange bearing
frog (#11)

Slide 10 of 19
“Rainy Section”
• Watering system and
contaminated ballast
 20-foot test section
 Moisture and contamination
thresholds for rapid track degradation

Slide 11 of 19
Ties and Fasteners

• 2015-2017: 860 mixed hardwood ties

installed
 Effects of plate size and hold down type
• 2015: 100 engineered polymer
composite (EPC) ties installed
• 2018: 100 engineered polymer
composite (EPC) ties installed
 Evaluating lateral resistance, spike
retention, potential development of
fatigue cracking – both types of EPC ties

Slide 12 of 19
Revenue Service Testing

• 2018 focus areas:

 Heat affected zone (HAZ) overlay
treated thermite welds (test ending 2019)
 Performance of frog systems
 Broken spikes
 Top of rail friction control
 New generation insulated joints
 Cold climate concrete tie remediation
 Crossing diamond foundation
 Rail performance

Slide 13 of 19
Thermite Weld Heat-Affected Zone Overlay

• Results at WMS different

than at FAST
 Five-degree curves at FAST – 240 MGT
- Reduced batter at HAZ, shell in weldment
(common failure mode at FAST)
 Tangent track at WMS – 515 MGT
- Shells in HAZ just outside overlay
- Untreated welds haven’t developed similar shells
- Overlay may contribute

Slide 14 of 19
Frog Systems

• Frog design
 Heavy point
 Heel design
 Platework design
• Frog casting weld repair methods
 Robotic arc weld
 Laser cladding

Slide 15 of 19
Broken Spike Research

• Broken spikes not uncommon at FAST and

higher-degree curves in revenue service
• Testing underway and planned
 Instrumented spikes – lab, FAST,
revenue service
- Effects of:
• Tie condition
• Location in a plate and spiking pattern
• Train braking
• Rail temps and rail neutral temperature

Slide 16 of 19
Top of Rail Friction Control in Warm and Cold Weather

• Three vendor products tested

• Effectiveness evaluated by
effect on lateral curving forces
• Statistical data analysis
completed
 Minor reduction in peak lateral
forces with TOR
 Non-test variables may have
affected results

Slide 17 of 19
Next Generation Insulated Joints

• Eight pairs of
joints from two
vendors
 No new failures
noted during
inspection
 Will continue to
monitor as
additional tonnage
accumulates

Slide 18 of 19
 Thank you!
Host railroads (UP, NS, CN)
U.S. Federal Railroad Administration
Transport Canada
TTCI Research Team
TTCI Instrumentation Team
TTCI Maintenance and Repair Teams

Slide 19 of 19
Slide 19 of 19
Emerging
Technologies
for Automated
Detection
Matthew W. Witte, Ph.D.

©2019 TTCI
TTCI is a wholly owned subsidiary
Slideof1 the AAR
of 21
TTCI Research for Automated Inspection and Detection

• TTCI’s SRI research

directed by industry
• Priorities set by
industry VPs
• Technologies and
methods of study
proposed by TTCI
• Focus on innovation
and implementation

Slide 2 of 21
What Technology Driven Train Inspection (TDTI)?

Wayside
The Technologies
Machine
Vision
Inspection

Wayside
Performance Vehicle
Detection DATABASE Health
Report

Wayside
Condition
Detection

Slide 3 of 21
Technology Driven Train Inspection – Industry Vision

• Enables automated train

condition assessment with
a diverse network of
wayside and onboard
technologies
• Continued safety
improvements by
expanding capabilities
 Dynamically identifies
potentially catastrophic defects
 Reduces accidents, incidents,
and unplanned service
interruptions
 Minimizes the need for static
visual inspections
Slide 4 of 21
Technology Driven Train Inspection

• Inspection
 Monitor condition
 Measure components
 Monitor performance
• Detection
 Validate that parts are OK
 Trend on measurements
 Identify defects and out-of-
spec conditions
• Performance Monitoring
 Measure performance
- Directly and Indirectly

Slide
Slide 55 of
of 21
 Wheels

• Critical defects form

from within
 WILD finds out of
round wheels
 WILD is good for limiting
damage to infrastructure
• UT for internal defects
 Automated Cracked
Wheel Detector Systems
(ACWDS)

Slide 6 of 21
Tycho ACWDS

• Outboard probes
added in 2018
• Finds subsurface
cracks that
precede VSR

Slide
Slide 77 of
of 21
Axles

Thermography finds cracks to surface

Slide 8 of 21
Truck Component Inspection

• Revenue service
evaluation of KLD
system
• FRA cooperative
research program
• System reliability
validated

Slide 9 of 21
Coupler Securement

• Machine vision
systems
• Coupler
securement

Coupler pin retainer plate is

securely bolted

Slide 10 of 21
Draft Components

Cross key
Cushion type

Slide 11 of 21
Rail Defects

• Head and web with Phased Array

Ultrasound (PAUT)
 Head contacting phased array
ultrasonic
 Compensates for profile wear
 Capable of finding TD under shell
• Base flange still a challenge
 2019 will focus on finding new
methods
Slide 12 of 21
Rail Defects
Welds and PAUT weld study
Lack of Fusion in Weld

Slide 13 of 21
Classifying Rail Defect Type

Neural Network Analysis

Fall-Spring Summer Winter
Model Model Model
Accuracy Accuracy Accuracy
Bolt hole crack
(BHC)
95.9 92.3 94.9

Vertical split
head (VSH)
83.7 84.2 78

Crushed head
(CH)
59.1 82 22.9

Slide 14 of 21
Rolling Contact Fatigue (RCF) on Rails

Electromagnetic Field Imaging (EMFI)

Contoured Sensor Non-contacting

Slide 15 of 21
EMFI Tests

• Walking
• Stick
• On-truck
• True depth

Slide 16 of 21
Thermal Signatures

LUTIS: Locomotive Undercarriage Thermal Inspection System

Single Camera LUTIS Dual Camera LUTIS

Slide 17 of 21
Thermal Signatures

• LUTIS
 Not just hot spot detection
 Trending thermal conditions

Slide 18 of 21
Data Analytics

Slide 19 of 21
Solution Network
Wayside Detector Network
• Trade organizations
(ASNT, IHHA, etc) for new
technologies
• Silicon Valley for
analytic techniques
• TTCI for innovation,
integration, development,
and testing
• RailInc for data handling
• Railroads’ back offices for
implementation
Slide 20 of 21
Thank you!

Slide 21 of 21
Slide 21 of 21
Track and Structure
Breakout Session

©2019 TTCI
TTCI is a wholly owned subsidiary
Slideof1the AAR
of 24
Rail Performance
and Integrity
Ananyo Banerjee, Ph.D.

©2019 TTCI
TTCI is a wholly owned subsidiary
Slideof1the AAR
of 12
Why Analyze Rail Performance and Integrity?
• Challenges for the Industry:
 Wear under heavy axle loads (HAL)
 Rolling contact fatigue (RCF)
 Fatigue defects
 Electric flash butt (EFB)
weld failures
 Wheel/rail lubrication
• End products:
 Innovate methods of testing
 Implement guidelines for rail
maintenance
 Improve life of rails

Slide 2 of 12
High Strength (HS) Rail Test (2014-2018): Wear and Grinding

• Test concluded at
651 MGT.
 Gage face wear
dominated
 Grinding:
- High and low rails
(255 and 429 MGT)
- Additional grinding
of low rails at
590 MGT

Slide 3 of 12
HS Rail Test (2014-2018): Weld Failures

Eight EFB weld failures on high rail involving all six rail types

Defect Initiation

Slide 4 of 12
HS Rail Test (2014-2018): RCF
RCF intensity variation: Comparison of crack progress with intensity

Mild-1 Heavy-2 Severe-3

0
Surface metal loss due to spalling

2.5 (0.1)

5.0 (0.2)
mm (in.)
Slide 5 of 12
New HS Rail Test (2018-now)

• Eight suppliers participating

• Rails have higher hardness
• Focus on EFB weld failures

20’ N A P E V S J B A E N S P V B J A S E N B P J V 20’

960 feet
Non-Test Plug Non-Test Plug
N NIPPON SUMITOMO A ARCELOR MITTAL E EVRAZ ROCKY MTN. B BRITISH STEEL
J JFE STEEL P PANZHIHUA ANGANG S STEEL DYNAMICS V VOESTALPINE SCHIENEN

Slide 6 of 12
Influence of Track Gage on Rail Performance (2016-present)

• Two 380-ft. zones of two track gages: 56 1/4 in. and 57 in.
• Over 300 MGT accumulated; one intermediate strength rail type used
• Head (vertical) metal loss on low rail more in 56 1/4 in. gage than 57 in. gage

Slide 7 of 12
Influence of Track Gage on Rail Performance (2016-present)

• Gage widening observed in both 56 1/4-in. and 57-in. zones

• Gage face wear on high rail slightly more in 56 1/4-in. gage zone

Slide 8 of 12
Influence of Track Gage on Rail Performance (2016-present)

• Four shells on high rail

• RCF on low rail:
 Narrow gage (56 1/4 in.) showing narrower RCF band
 Wide gage (57 in.) showing broader RCF band and sporadic spalls

Narrow RCF
band in 56 1/4″ Sporadic spalls
gage zone in 57″ gage zone

Slide 9 of 12
 Fatigue Defects under Simulated Heavy Axle Loads

• Testing at Texas A&M University

 Rails with internal defects subjected to alternating loads simulating
bending stresses
- First test lasted 1.3 million cycles~36.5 MGT
- Second test lasted 6.7 million cycles~191.6 MGT
 Thermal load simulated by longitudinal tensile stress

First Test Second Test

Slide 10 of 12
Outcomes and Upcoming Future of Rail Research
• Summary
 Gage wear dominated performance of HS rails in last test
- New test with harder HS rails started with focus on EFB weld failures
 RCF and wear differences in IS rail due to track gage variation
 Rail integrity evaluation by new methods
- Neutron diffraction provides detailed residual stress measurements
- Simulated loading show huge variation in growth of fatigue defects
• Future Work
 Investigation of rail base corrosion under HAL
 Lubrication effects on RCF ― a broader study
 Further investigation of the complex relationship between residual,
bending, thermal stresses and internal fatigue defects
Slide 11 of 12
 Thank you!
Acknowledgements:
Yuqing Zeng, Kenny Morrison, Greg Giebel - TTCI
Rail Manufacturers
AAR Member Railroads

Slide 12 of 12
Slide 12 of 12
Improved Rail
Inspection
Technologies
Matthew Witte, Ph.D.
Anish Poudel, Ph.D.

©2019 TTCI
TTCI is a wholly owned subsidiary of the
Slide 1 ofAAR
13
 Improving Rail Inspection Technologies
Head

• Avoid Rail Breaks Web

• Inspect the entire section for defects Base

 Defect detection over all conditions: Rail Cross Section

2018 Focus - RCF

- Base Flange Defects
- Grade Crossing
- Weld Inspection
 Speed and detection efficiency RCF Damage

• End Products
 Facilitate the development of validated next
generation rail inspection systems

Base Defects Slide 2 of 13

Rail Rolling Contact Fatigue
The Challenge: Focus:
 Rolling contact damage  True measurement of:
- Fatigue cracks and pits - Crack depth
 Complicates rail inspection - Pit depth
- Remediated by rail grinding
 Subjective scale: 1, 2, or 3

Mild (1) Heavy (2) Severe (3)

Slide 3 of 13
Athena EMFI Technology – True Crack Depth
Athena Industrial Services EMFI

Air Gap

Contoured Sensor Channel Locations Non-contacting

Slide 4 of 13
Athena Echo-3D Data

• 3D depth map
 True crack and
pit depth
 Digital data
 Map and
monitor RCF

Slide 5 of 13
Athena EMFI Technology Testing

EMFI sensor

Hand-pushed cart with contoured EMFI sensor On-track testing with high-rail truck

Slide 6 of 13
EMFI Accuracy Test
Grind Profiles
Grind 5

Grind 4

Grind 3

Grind 2

Grind 1

Pre Grind
No Grind Grind zone

Slide 7 of 13
EMFI Repeatability Test

Slide 8 of 13
Phased Array (PAUT) Rail Weld Study

• Determine the effectiveness of hand-held

PAUT technology for characterizing rail
weld defects
 Detecting defects within welds
 Sizing defects
 Evaluating material
condition

Slide 9 of 13
Weld Study Thrust

• Establish PAUT sensitivity for

characterizing volumetric
rail weld defects:
Indication for
 Porosity weld with low
level porosity
 Slag inclusions
 Lack of fusion
 Shrinkage
• Recommendations for
weld inspection

Slide 10 of 13
 Lack of Fusion Weld Result

Defect Location

Lack of Fusion Sizing:

0.3 in. height by 0.4 in. wide at 2.1 in. deep
Hand-held scan of weld

Slide 11 of 13
Classifying Rail Defect Type

Neural Network Analysis

Fall-Spring Summer Winter
Model Model Model
Accuracy Accuracy Accuracy

Bolt hole crack

(BHC)
95.9 92.3 94.9

Vertical split head

(VSH)
83.7 84.2 78

Crushed head
(CH)
59.1 82 22.9

Slide 12 of 13
 Thank you!
Athena Industrial Systems, Alberta, Canada
TTCI team:
Anish Poudel, Brian Lindeman,
Branden Lawson, and Abe Meddah

Slide 13 13
Slide of 13
of 13
Bridges
Duane Otter, PhD, PE
Anna Rakoczy, PhD

©2019 TTCI
TTCI is a wholly owned subsidiary
Slideof1the AAR
of 14
Bridges

• Challenges addressed:
 Increased loading
on bridges
 Increased traffic
volumes and train speeds
 Alternative technologies
and materials
• End products:
 Longer safe service life
for bridges

Slide 2 of 14
Railroad Bridge Statistics

• About 100,000 railway bridges in the U.S.

(vs. 600,000 highway bridges)
• Over 2,000 miles of railway bridges
• Over $100 billion replacement value

Slide 3 of 14
Slide 3 of 14
Railroad Bridge Statistics

• Materials
 53% steel
 23% concrete
 24% timber
• Ages
 Minimum design life 80 years
 Median life about 100 years
 Most were designed for
heavy steam locomotives

Slide 4 of 14
Bridge Research: 2018 Focus Areas
• Fitness for Service
Assessment for Steel
Spans
• Analytical Support –
Bridge Life Software
• Effects of Double
Stack Traffic
• Next Generation
Bridge Decks
• Revenue Service
Implementation

Slide 5 of 14
Steel Bridge Fitness for Service (FFS): Bridge Life Testing

• Riveted deck plate girder (DPG) spans

• Five DPG spans dating to 1904 at FAST
• Complemented by revenue service tests
on NS and BNSF

Slide 6 of 14
Steel Bridge FFS: Bridge Life Estimation

• Spans are heavily loaded

by FAST train
• Data being used to develop
FFS methodology
• Implementation in progress
with AREMA 15
• Significant improvements in
life estimates (often more
than 200 years)

Slide 7 of 14
Steel Bridge FFS: Bridge Life Estimation

• For most spans longer than

car length – only one cycle
per train
 Stress ranges < 6 ksi
• Average impact much lower
than design impact (for rare
events)
• Effects from deck,
bearing friction
• Life based on cube of stress
range

Slide 8 of 14
Steel Bridge FFS: Bridge Software

• New methods not easily

calculated in spreadsheet
• Software being developed
to assist railroad bridge
engineers with FFS calculations
• Beta test version with railroads

Slide 9 of 14
Bridge Loading: Effects of Double Stack Cars

• Analytical study
completed
• Short spans and
floor systems identified
as most affected
• Open deck ties also
a concern

Slide 10 of 14
Slide 10 of 14
Bridge Loading: Effects of Double Stack Cars

• Testing completed on short span bridge with double stacks and

coal trains
• Coal trains have all cars fully loaded
• “Pillows or pig iron” in stacks (often not loaded to full weight)
• Intermodal train results in less fatigue accumulation
12 12
Intermodal
SS_mid Train
NS_mid CoalSS_mid
Train NS_mid
10 10

8 8
Stress, ksi

Stress, ksi
6 6

4 4

2 2
Time, sec Time, sec
0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120 140
-2 -2

Slide 11 of 14
Next Generation Bridge Decks

• Alternative Ties
 Glued-laminated timber ties (1,287 MGT)
 Fiber-reinforced foamed urethane
(fiberglass) ties (1,107 MGT)
 Superelevated glued-laminated ties
(experimental first trial) (159 MGT)
• Alternative Deck Fasteners
 Between-the-ties hook bolts – safer,
faster installation (307 MGT & 159 MGT)

Slide 12 of 14
Bridge Research

• 2018 Progress:
 FFS assessment showing better
bridge life estimates
 Bridge life software in beta-test
 Double stack traffic: watch
floor systems and deck ties
 Next generation bridge decks ―
promising alternative ties and
deck fastening system

Photo courtesy of Norfolk Southern Railroad Slide 13 of 14

Slide 13 of 14
Thanks to:
AAR Bridge TAG
Steve Dick
David Linkowski

Photo courtesy of Norfolk

Southern Railroad

Additional photos
courtesy of:
Canadian Pacific Railroad
Norfolk Southern Railroad

Thank you!
Joe Blackwell
Nathan Zachman
Photo courtesy of Norfolk Southern Railroad
Slide 14 of 14
Slide 14 of 14
Ties and
Fasteners
Yin Gao, Ph.D.

©2019 TTCI
TTCI is a wholly owned subsidiary
Slideof1the AAR
of 14
Improved Tie and Fastener Systems
• Challenges addressed:
 Economically improving track
performance – improvements of
the tie/fastener system
 Understanding where and why
tie/fastener failures occur
Composite Ties
• End products:
 Recommendations for
implementing improved designs
 Improved modeling for better
maintenance/purchasing decisions

Concrete Ties Wood/Composite Ties

Slide 2 of 14
Composite Tie Program
• Engineered Polymer Ties (EPC)
• In-track Tests at FAST and Revenue Service
 Some EPC ties have lasted over 2.1 billion gross tons
FAST 6-degree curve HAL HAL western mega site Chester, IL

Slide 3 of 14
Composite Tie Program

• Three failure modes:

Spike hole cracking
 Tie plate cracking
 Spike hole cracking
 Tie center cracking
• Qualification AREMA Tie plate cracking
lab testing
 Rail/plate area
compression
 Spike insert pullout
 Center negative bending Center cracking

Slide 4 of 14
Composite Tie Program

• Fatigue Test Development

 Four-point bending: Address center cracking
 Modified AREMA Test 6, wear/abrasion: Address spike hole cracking

Slide 5 of 14
Composite Tie Fatigue Bending Test
• Composite ties fail differently than other tie types
• Voids, inclusions, discontinuities, weak areas – potential failures

Slide 6 of 14
Composite Tie Fatigue Bending Test
• Constant moment/bending stress is as large P
an area as possible
• Increased load to increase the
severity of the test
• Test recommendation:1.5 M cycles at 5Hz
15" 30" 15"
L

Slide 7 of 14
Pressure at the Tie-Ballast Interface
• Composite ties are generally less stiff than wood ties
• Geotrack modeling predicts higher pressure at the tie-ballast
interface under the railseat
• May cause a center-bound condition more quickly

Slide 8 of 14
Thermal Influence on Composite Ties
Tie end moves due to bending
8 AM: Tie Temperature 50°F
3 PM: Top of tie temperature 90 °F

Slide 9 of 14
Wood Tie Test Zone

• Wood tie test zone Tie Plate

14″ AREMA Rolled
Rail Fastener
Cut Spike
Hold-Down
Cut Spike
installed in 2017 summer 16″ AREMA Rolled Cut Spike Cut Spike
(800 ties) 18″ AREMA Rolled Cut Spike Cut Spike
16″ Pandrol Victor Rolled eClip Cut Spike
 Focus on comparing plate 16″ Pandrol Victor Rolled eClip Coach Screw
size/type and spike types 16″ Pandrol Victor Rolled eClip Drive Spike
(200 MGT so far) 18″ Pandrol Victor Rolled eClip Cut Spike
18″ Pandrol Victor Rolled eClip Coach Screw
18″ Pandrol Victor Rolled eClip Drive Spike
Slide 10 of 14
 Spike Breakage Study
• In general: Broken spikes found mostly on elastic fastener tie plates
 Historically observed in premium fastener systems at FAST and
in revenue service
- High rail: more often on steep grade, high degree curves
- Possible contributing factor in recent derailments in revenue service

Slide 11 of 14
Spike Breakage Study
• Uplift of elastic fastener tie plates
• Modeling of the stresses in spikes

Slide 12 of 14
Spike Breakage Study
• Four surfaces instrumented with strain gauges
• Instrumented locations determined by field observation
 Rail spike: 3 1/4 in. below the bottom of head
 Anchor Spike: 2 1/2 in. below the bottom of head
• Calibration and in-track tests

Slide 13 of 14
 Thank you!
Acknowledgements
AAR Member Railroads
Federal Railroad Administration
AREMA Committee 30
Composite Tie Manufacturers
TTCI Research Team

Slide 14 of 14
Slide 14 of 14
Substructure
Systems
Stephen Wilk, Ph.D.

©2019 TTCI
TTCI is a wholly owned subsidiary of the
Slide 1 of AAR
18
Substructure Systems

• Challenges:
 Inspection methods
 Ballast characterization and performance
 Ballast maintenance
 Subgrade remediation
• End products:
 Improve safety
 Improve maintenance
effectiveness
 Improve maintenance
planning

Slide22ofof18
Slide 18
Substructure Systems – 2018 Research

Research Project Challenge Research Topic

Ballast characterization Effect of moisture in fine-contaminated
Rainy section
and performance ballast
Rainy section Ballast maintenance Effect of lift height during spot tamping
Reducing surface maintenance at ballast
Captina site Subgrade remediation
pocket location
Reducing subgrade pumping at flooded track
Cleveland site Subgrade remediation
zone

Slide 3 of 18
Rainy Section - Background
• Fine-contaminated ballast Rainy
Section
 Common issue with railroads
 Multiple variables affect performance
- Moisture
- Fine level, fine size, fine plasticity
• Large-scale “laboratory” test
• Research scope:
 Effect of moisture on
fine-contaminated ballast
 Changes in drainage condition
from mud pumping
 Benefits of maintenance activities

Slide 4 of 18
Rainy Section – Background
• Located in Section 36 of
High Tonnage Loop (HTL)
• 20-foot section
• 40% fines
 Most ballast voids filled
 Fines from natural
ballast degradation
• Irrigation and
drainage system
• FAST train
 ~2 MGT per night
of operation
 40 mph

Slide 5 of 18
Rainy Section – Effect of Moisture

• Settlement rate three to six times greater when wet

• Similar dry pre- and post-mud pumping behavior
Settlement with Distance Settlement Rate – Rain Events

Slide 6 of 18
Rainy Section – Effect of Moisture
• Increased displacement during mud pumping
• Decreased track modulus when wet and mud pumped
Rail and Tie Displacements Track Modulus

Slide 7 of 18
Rainy Section – Mud Pumping Mechanisms
• High moisture levels around tie end
 Ponding around tie
 Pumping rearranges fines near surface
• Other mechanisms may exist
 Flooded cut regions (Cleveland site)

Slide 8 of 18
Rainy Section – Mud Pumping Mechanisms
• Preliminary relation between moisture and settlement rate
 Crib sensor used
• 15% moisture – fines able to flow (site specific!)
 Related to surface mud pumping
• No correlation with deeper moisture sensors

Slide 9 of 18
Rainy Section – Drainage

• Mud pumping inhibits surface drainage

 Clean ballast: minutes
 Pre-mud pumping: ~1 day
 Post-mud pumping: ~5 days
• Increased “softened” time
 More susceptible to track
geometry degradation
• Future Work – Maintenance
 Breaking up surface
 Trenches
 Shoulder cleaning

Slide 10 of 18
Rainy Section – Spot Tamping
• Initial settlement after tamping
 Ballast consolidation phase
 Majority in first 0.1 MGT

Slide 11 of 18
Rainy Section – Spot Tamping
Lift Material Wet or
Test
• Higher lift result in longer Height Added? Dry?

lasting track geometry 1 0.52 in. No Dry

2 0.51 in. Yes Dry
Test 3 – 1.47-inch lift 3 1.47 in. Yes Dry
4 0.07 in. Yes Wet

Initial settlement / lift height

Slide 12 of 18
Captina Site – Background

• Embankment with ballast pockets

 6 to 8 feet deep
 Soft embankment material
• Line history
 Norfolk Southern line
 Operations for 100+ years
 ~17 MGT line
 10 mph
• Weekly surface maintenance

Slide 13 of 18
Captina – Geogrid/Ballast Drain Remediation

• 2012 Remediation
 Geogrid
 Ballast drains
• Significant reduction
in maintenance
 Weekly to yearly to quarterly
• Payback period = ~1 year
• Lower embankment issues

Slide 14 of 18
Cleveland Site – Background
• Low region that experiences Bottom of tie
flooding
• Line history
 Norfolk Southern Line
 ~20 to 25 MGT line
• Subgrade pumping
into ballast
 Flooding
 Broken up shale
 Drainage difficult

Slide 15 of 18
Cleveland Site – Geosynthetic Remediation

• Geosynthetic remediation (2016)

 Barrier layer to prevent
subgrade pumping
 Three geosynthetics tested
• Four ballast sampling
locations (2018)
• Benefits of geosynthetics
 Reduced fines in ballast layer
 Behaved similarly

Geogrid bonded Tracktex Geocell with

w/fabric geofabric base
Slide 16 of 18
Continuing Work

• Rainy Section
 Moisture effects
 Maintenance practices to
improve surface drainage
 Spot tamping
• Production tamping
• Ballast pocket locations
 Inspection techniques
• Substructure
Management System
• Ground penetrating radar
(GPR) and track
geometry degradation
Slide 17 of 18
 Thank you!
Acknowledgements:

Colin Basye and Yin Gao (TTCI)

Federal Railroad Administration

Slide 18
Slide 18 of
of 18
18
Implementation of
Improved Special
Trackwork
Foundations
Benjamin Bakkum, P.E.

©2019 TTCI
TTCI is a wholly owned subsidiarySlide
of the AAR
1 of 14
Special Trackwork

• Challenges of special trackwork

 Components wear and fail prematurely
 Only remediation options in most
cases are:
- Condition based speed restriction
- Component replacement/repair
 Special trackwork component
lifespans relatively short when
compared with other areas

Slide 2 of 14
Over 20 Years of Crossing Diamond Research
• Innovate:
 Flange bearing diamond developed – decreases
dynamic impact loads
 Dampening pads for under ties/platework/casting
developed to further dampen impact loading and
maintain surface conditions
• Implement:
 Crossing diamonds now more commonly installed
with dampening pads
• Improve:
 Current focus is optimization of dampening pad
layers, thickness, and material properties

Slide
Slide 33 of
of 14
Special Trackwork Foundations (con’t.)

• Initial study on dampening pads

at FAST
 Showed more uniform track stiffness
 Showed less overall settlement
• Only one layer of dampening pads
 Located under ties and approaches in test
panel case
• Other tests at FAST:
 Diamond with under-casting pads

Slide 4 of 14
Switch 408 at FAST
• No. 20 turnout installed in 2013
• Installed with under tie pads throughout. Since install:
 Out of face surfacing in 2014
 500+ MGT

Slide 5 of 14
 Case Study #1

• Existing diamond
 Two pad layers
- Under-tie and under-platework
 Conventional tread bearing
• Replacement diamond
 Three pad layers
- Under-tie, under-platework,
AND under-casting with
milled platework
 Also conventional tread bearing

Slide 6 of 14
Current Performance- Case Study #1

• SW corner Acceleration Range by Location

 Accelerations observed similar to 200
150
other previous tests 100

• NE corner 50

Acceleration [g]
0

 large increase in observed on -50

-100
platework and casting -150

 Likely cause is air gap between

-200
-250

under-platework pad and platework -300

Case 1 - SW Case 1 - NE
Casting Plate Tie

Slide 7 of 14
Air Gap Analysis
SW CORNER

NE CORNER

Slide 8 of 14
NE Corner Vertical Movement

Slide 9 of 14
Case Study #2

• Existing Diamond:
 Tread-bearing
 No pads
• Replacement
Diamond:
 OWLS:
One-Way-Low-Speed
 Under-tie pads
 To be installed
in 2019

Slide 10 of 14
Comparison of Current Performance

• Case Study 2 site

shows higher
accelerations
 Potential cause: no pads
 Case Study 2 site also
had higher traffic speed

Slide 11 of 14
Continuing Work

• Cold weather analysis of Case Study 2 site-

 Monitor this site in both frozen and unfrozen conditions to study
effect of climate on pad performance
• Comparison of original vs. replacement diamonds
 Case Study 1- impact of additional pad layer vs. original

Slide 12 of 14
 Other Continuing Work

• Diamond
Foundation Tests
 Two additional
locations in Ohio
 Evaluating pad
performance and
material properties

Slide 13 of 14
 Thank you!
TTCI: Stephen Wilk, Dave Davis, Duane Otter
AAR Member Railroads
Transport Canada

Slide 1414
Slide of of
1414
Mechanical
Breakout Session

©2019 TTCI
TTCI is a wholly owned subsidiary
Slideof1the AAR
of 24
Car and
Truck Systems
Russell Walker

Slide 1 of 13
 Truck Systems

• Safety
 Good truck performance
reduces derailments
• Economics
 Good truck
performance reduces:
- Truck maintenance
- Wheel replacements
- Fuel use
- Rail wear
• Overview
 Million mile truck teardowns
 Accident/incident
database analysis
 Centerplate chamfer

Slide 2 of 13
Analysis of Worn Truck Inspection and Test Data

• Detailed inspection of 120%

four car sets

Wedge Rise (percent of limit)

100%

• The test trucks were in 80%

overall good condition A B AAR Limit
60%

40% Trucks selected for

detailed inspection are
marked with an ×
20%

0%
0 250,000 500,000 750,000 1,000,000
Mileage

Slide 3 of 13
Analysis of Worn Truck Inspection and Test Data

Some individual truck components were worn beyond limits

Bolster pockets
Column
wear plate

Nine broken inner springs,

one inner control coil
Pedestal Thrust Lug
Slide 4 of 13
Analysis of Worn Truck Inspection and Test Data

• Maintenance data indicates

 Nine adapter pad replacements (all on same car,
different style pad)
 One stabilizing spring replacement
 Twenty-seven wheelset replacements
• Reasons for wheelset removals
 Twelve high impact wheels
 Seven bearing-related removals
 Six thin flanges
 Two high flanges
• Truck performance influences removals
due to high impact wheel, thin flange,
or high flange
Slide 5 of 13
Analysis of Worn Truck Inspection and Test Data

• Tested examples of worn M-976

trucks
 Empty Hunting – stable to about 60 mph
 Loaded Hunting – stable to about 50 mph
 Loaded dynamic curve
- Range of speeds <18 mph where wheel lift
observed
- One truck showed high L/V ratio even with
acceptable vertical wheel load
 Loaded twist and roll – no wheel lift, but less
than 10% of static load at resonance
 Loaded pitch and bounce – spring bottoming
at about 50 mph
• M-976 trucks don’t meet criteria
for a new truck, but they don’t
have a reputation as bad actors
Slide 6 of 13
Analysis of FRA Accident/Incident Database

• Derailments occurring
between 2010 and 2017 5

Cumulative Cars Built

Since 2003 (100,000)
• Mechanical cause codes 4

• Queried UMLER with 3

first car derailed 2
 Build date 2003 or later 1
 Increased Gross Rail Load 0
code to identify M-976 and

2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
non M-976 trucks
Year
M-976 Non M-976

Slide 7 of 13
Analysis of FRA Accident/Incident Database

• M-976 fleet appears to perform better

5

cars built since 2003

4
than non-M-976 fleet

Derailments of
3

• Most common cause is truck 2

bolster stiff, improper swiveling followed 1

by truck bolster stiff (failure to slew) and 0

truck hunting 2010 2011 2012 2013 2014 2015 2016 2017
Year
Non M- Non M-976 M-976
Cause Description M-976
976
E46C Truck bolster stiff, improper swiveling 13 2 * Includes derailments for all track
E47C Defective snubbing (including friction and hydraulic) 1 0 types ( Main, Yard, Siding and
Other truck component defects, including Industry)
E49C 2 0
mismatched side frames
E4BC Truck bolster stiff (failure to slew) 7 1
E4TC Truck hunting 2 1
Total 25 4

Slide 8 of 13
 Centerplates
• Reducing stress by
increasing radius at
chamfer
• Testing four plates at FAST
60,000
Extrusion
Maximum Pressure (psi)

50,000

40,000

30,000

20,000
Standard Center Plate Modified Center Plate
10,000
3/8 in. 20 in.
0
0 10 20 30 40
Radius at Chamfer (inch) Truck Bolster 1 in. Truck Bolster

Slide 9 of 13
Original Center Plate at One Week
Line of contact
already visible

Slide 10 of 13
Modified Center Plate at One Week
No line of contact,
but some extrusion
already visible

Slide 11 of 13
Summary

• Inspections and tests

 M-976 trucks in good condition
after nearly one million miles
 Do not meet test requirements
for new trucks
• Accident/incident analysis
 M-976 seems to perform well
compared to other trucks
 Truck turning is leading cause
• Modified centerplate
 Under test at FAST

Slide 12 of 13
 Thank You!
Acknowledgements:
BNSF
Mitsui Rail Capital
Amsted Rail
Standard Car Truck
Lewis Bolt and Nut

Slide
Slide 1313
of of
1313
Impact Modeling
of Cars Equipped
with End-of-Car
Cushioning Units
Adam Klopp
Scott Cummings
Jack Schultz
Stan Gurule

©2019 TTCI
TTCI is a wholly owned subsidiary of the
Slide 1 ofAAR
13
Load Environment of Trains

• Challenges Addressed
 Derailment and/or track damage
from run-in events
 Damage to rolling stock and/or lading
 Train separation
• End Products
 Load environment for cars equipped
with cushioning units
 Recommendations for draft system
improvements based on simulations

Slide 2 of 13
End-of-Car Cushioning (EOCC) Units

• Type of hydraulic draft system

 Absorb energy by forcing oil from
the high pressure cylinder to the
surrounding casing
 Velocity dependent resistance
 Long travel
- Typically 10 or 15 inches per unit
• Designed to protect lading
during coupling events
 Provide necessary protection in yard
environments
 Can also create slack action issues
during normal operations
Slide 3 of 13
Modeling of Cars with EOCC Units

• Models developed in NUCARS and TOES NUCARS

to evaluate EOCC units and other draft systems
 NUCARS: Yard coupling events between two cars
 TOES™: Over-the-road operations with full train
• Initial efforts focused on modeling 15-inch
EOCC units
 Based on available reference and test data
 Simulate to M-921B standard EOCC Unit

Slide 4 of 13
NUCARS Impact Model Description

• NUCARS model was developed with four instances of the

standard loaded hopper model
 Represent the hammer, anvil, and two string cars
- Weights adjusted according to M-921B standard
 Each model modified to add couplers and draft systems
 Active representation of the wheel-rail interface
 Includes the normal degrees of freedom for full vehicle models
Z (Vertical)

Y (Lateral) Yaw

X (Longitudinal)

Pitch Roll
Slide 5 of 13
NUCARS Impact Model

String Car 2 String Car 1 Anvil Hammer

Full NUCARS Model

Slide 6 of 13
Animation of M-921B Impact in NUCARS

Slide 7 of 13
Comparison of NUCARS and Measured Impact Data

• NUCARS results of 15-inch EOCC 300

Peak Coupler Force (kips)

0
unit impact correlated well with -300
measured data -600

 Both buff and draft impacts -900

0 1 23 4 5 6 7 8 9 10 11
 Multiple supplier units Impact Speed (mph)
NUCARS Draft Force Measured Draft Force
• TOES model also constructed NUCARS Buff Force Measured Buff Force

to simulate M-921B type impacts 5

Displacement (inches)
 Comparison with NUCARS 0

Peak Coupler
-5
 Verify same characteristics -10
 Correlated well with NUCARS -15
up to 8 mph impact 0 1 23 4 5 6 7 8 9 10 11
Impact Speed (mph)
NUCARS Draft Disp. Measured Draft Disp.
NUCARS Buff Disp. Measured Buff Disp.

Slide 8 of 13
 TOES Over-the-Road Simulations
• Progressed from impact simulations in NUCARS to over-the-
road (OTR) simulations in TOES
• Challenging revenue service routes
 Undulating grades, hill crest operation, sag negotiation
 Based on actual operating scenarios
- “15-inch” coil steel unit trains, “10-inch” autorack unit trains

NUCARS 
Impact

TOES
OTR

Slide 9 of 13
TOES Run-in Simulation: Unit Coil Steel Trains
Baseline Route with Run-in event

Peak Buff Force:

236 kips at Car
26

Slide 10 of 13
TOES Run-in Simulation: Unit Coil Steel Trains
Baseline – Route with Run-out
Peak Draft Force:
185 kips at Car 48

Slide 11 of 13
Future Work
10-inch EOCC
• Simulation effort still ongoing
• Modeling other draft systems
planned for 2019
 Finish 10-inch EOCC modeling
 Active draft EOCC
 EOCC with other pre-loads
 Long travel draft gears
• Research will help to isolate and
identify characteristics for EOCC with Active Draft

improved draft system

performance

Slide 12 of 13
 Thank you!
BNSF Railway
Norfolk Southern Railway
Canadian National Railway
TTCI Staff

Slide 13 of 13
Slide 13 of 13
Brake Systems
Scott Cummings

©2019 TTCI
TTCI is a wholly owned subsidiary
Slideof1the AAR
of 14
Brake Systems
• Challenges addressed:
 Undesired Emergency Brake Applications (UDEs)
 Component life extension
• End products:
 Reduced line-of-road failures
 Recommended repair practices
 Consistent component behavior
• 2018 focus areas:
 UDEs: control valve stability
 Effect of tread braking on wheel life
 Performance of tread conditioning brake shoes

Slide 2 of 14
Brake Systems
• UDE: Brake control valve stability
 Sixty emergency valve portions categorized by
- Condition
• New or reconditioned
• Suspect bad actor
o No obvious problem (NOP)
o Failed single car air brake test (SCABT)
o Trainline issue
- Model
• Older and unstabilized portions
• Newer portions
 Define stability threshold for each valve
- Laboratory short duration brake pipe pressure fluctuations
Slide 3 of 14
Brake Systems

• UDE: Brake control valve stability

 SCABT and brake system
inspection are critical tools Green = Most Stable

Stability Threshold (psi/sec)

Blue = In-between
 Least stable valves Red = Least Stable

- SCABT and/or
trainline issues
- Older, unstabilized portions

Valve Types
1&2 Older, Un-stabilized
3 Older, Stabilized
4&5 Newer, Short Car Valve Type:
Valve Condition:
6&7 Newer, Long Car

Slide 4 of 14
Brake Systems
• Effect of tread braking on wheel life Shells found on a wheel with no functioning brakes

 Enforcement discretion for disabled

brakes in service
 Three articulated double stack cars
- Wheelset removals by 150,000 miles
- Five of six in test group
- Four of six in control group
Test Control
Truck Truck

Slide 5 of 14
Brake Systems
• Effect of tread braking Typical shelling band on wheel tread
on wheel life:
 Ten coal hoppers
(five with disabled brakes)
- No removals due to
wheel condition as
of 110,000 miles
- Inspection shows minor
tread issues
• 37% of test group wheels
• 16% of control wheels

Slide 6 of 14
Brake Systems

• Effect of tread braking

on wheel life
 Ten coal hoppers (five with
disabled brakes)
- Wheel wear comparisons
show no advantage
to disabled brakes

Flange Wear
Tread Wear

Slide 7 of 14
Brake Systems
• Tread conditioning (TC) brake shoes
 Promoted as a way to provide cost savings in
both wheel and brake shoe replacements
• Survey of six revenue service tests Three types of
shoes tested
 Most reported improvements in wheelset life
 One reported improvement in shoe life
• Laboratory dynamometer tests
 Nine shoes tested
- Three TC “A” Brake
Dynamometer
- Three TC “B”
- Three high friction composition (HFC)

Slide 8 of 14
Brake Systems
• Dynamometer testing
 Wheels Flat spots 3 in. long,
0.075 in. deep
- Freshly trued
- Machined flat spots
- Cold worked
 Eight cycles of AAR M-997 per wheel/shoe

Test type Quantity Duration Speed Shoe force

Light grade 1 45 minutes 20 mph 1,450 lbs.
Light stops 6 Until stopped 40 and 60 mph 1,500 lbs.
Heavy stops 6 Until stopped 60 and 80 mph 6,020 lbs.
Heavy grade 1 45 minutes 20 mph 2,250 lbs.

Slide 9 of 14
Brake Systems

Groove worn
during testing
• Dynamometer testing
 Each shoe matched
Flat spot with a different wheel
 Each shoe produced
measurable wear in
line with the flat spot,
effectively reducing
Wheel always rotates
this direction
the runout

Slide 10 of 14
Brake Systems
Zoom view of wheel tread Initial profile at
flat spot
Initial profile 6 inches
from flat spot

Initial and final runout

Final profile 6 inches Profiles

from flat spot captured
every
6 inches

Slide 11 of 14
Brake Systems

• Performance of TC brake shoes

 Brake shoe wear
- Higher for “A” type tread
HFC TC A TC B
conditioning shoes
 No statistical difference
- Nominal wheel wear
- Local radial runout reduction
 Dyno test results versus
service experiences
 Coefficient of friction considerations
HFC TC A TC B

Slide 12 of 14
Brake Systems
• Undesired emergency brake applications
 SCABT and visual brake system inspection are good tools
 Least stable valves:
- Older models
- On cars that failed SCABT or had a trainline issue
• Tread braking
 Not essential for wheel tread degradation
 Laboratory tests did not show benefits for TC shoes:
- Shoe life
- Wheel runout reduction

Slide 13 of 14
 Thank you!
BNSF Railway
Norfolk Southern Railway
Canadian National Railway
TTX
Southern Company
Railinc
TTCI: Nick Hudnall, Mitch Miller,
Tony Sultana, Nick McLaren, and Bea Rael
Slide 14 of 14
High Performance
Wheel Research
Kerry Jones

©2019 TTCI
TTCI is a wholly owned subsidiary of 1the
Slide of AAR
16
High Performance Wheels
• Challenges addressed:
 Wheel damage
- Shelling and high impact
- Vertical split rims (VSRs)
• End products:
 Increased wheel life
 Reduction of VSRs
• 2018 research:
 High Performance Wheel Test 1 (HPW1)
- Revenue service test results
 High Performance Wheel Test 2 (HPW2)
- Durability test at FAST
Slide 2 of 16
Types of Wheel Damage

• Surface-initiated cracks
 Easily monitored by visual inspection

• Subsurface cracks
 Surface damage not always present
 Can develop into catastrophic
wheel failure
 Detectable by ultrasonic methods

Slide 3 of 16
High Performance Wheel Test 1 (HPW1)
• Tested wheels that offered better performance than Class C
• Head-to-head comparison: HPWs and Class C wheels

Slide 4 of 16
High Performance Wheel Test 1 (HPW1)

• Laboratory testing • Revenue service test

 Mechanical – tensile, yield  Western U.S.
 Metallurgical – cleanliness,  Initially Seven HPW types
structure  Four HPW types remain
• Durability testing at FAST  Class C – 304,000 miles
 Ran for 20,000+ miles before  HPW – 327,000 miles
beginning revenue service

Slide 5 of 16
 What Factors Make the Best Wheel ?

• Very complex issue

• No single parameter
controls
performance
• Many can influence
• No definitive answer

Slide 6 of 16
HPW1 Revenue Service Results – Top Three Wheels

Points show
wheels
removed for
shelling or
high impact
Two early
HPW removals

HPW removals
shifted, indicating
increased life

Removal
rates
similar

Slide 7 of 16
High Performance Wheel Test 2 (HPW2)

• Objective: Tighten wheel

specification to correlate
with performance
Test structure same as HPW1
 Laboratory testing – complete
 Durability testing at FAST – in progress
 Revenue service testing – begin 2019
• Fourteen steels, 11 manufacturers

Slide 8 of 16
Microcleanliness
• Measures discontinuities in 200+ fields per analysis
• Depths ~ 0.080 to 0.800 inch
Field

Analysis
planes

Void

Discontinuity
Sulfide
inclusion content 0.23%

Slide 9 of 16
HPW2 Laboratory Test Results

• Tensile, yield strength

• Fracture toughness
• Microstructure

Slide 10 of 16
HPW2 – Durability Test

• Forty wheelsets under

10 newer, similar cars
• Average 24,000 miles
• Minimum 25,000 miles
before revenue service

Slide
Slide 11
11 of
of 16
Wear of Test Wheels at FAST

• Wheelsets have
travelled an average
of 24,000 miles
• Wear data is for
informational
purposes only
 Small sample size
 Track geometry can
lead to accelerated
wear values

Slide 12 of 16
HPW2 Wheel Tread Condition (FAST)

Pitting

Surface-initiated cracks

Slide 13 of 16
HPW2 Revenue Service Test
Photo courtesy of up.com

• Hosted by Union
Pacific Railroad
• Install under grain cars
• Compare Class C and HPWs
 25 wheelsets per supplier
 400+ wheelsets total
• Routine visual and periodic
ultrasonic testing

Slide
Slide 14
14 of
of 16
Summary

• HPW1
 HPWs generally showing improved life over Class C wheels
- Removal rates similar, but HPWs failing later
• HPW2
 Wheels have travelled about 24,000 miles at FAST
 Four wheelsets removed for large subsurface cracks
 UP will host revenue portion of test, beginning 2019

Slide 15 of 16
Thank you!

Slide 16 of 16
Slide 16 of 16
Wheel Profile:
Design and
Maintenance
Scott Cummings

©2019 TTCI
TTCI is a wholly owned subsidiary of the
Slide 1 ofAAR
12
Wheel Profile: Design and Maintenance

• Challenges addressed:
 Wheel life – wear, fatigue
 Rolling resistance
 Safety – wheel flange climb
• End products: Wheel Profile
 Optimized wheel profiles
 Recommended maintenance
practices
• 2018 focus areas:
 AAR-2A wheel profile
 Optimal wheel rim thickness
 Hollow wheel wear limits

Slide 2 of 12
Wheel Profile: Design and Maintenance

• AAR-2A wheel profile for freight cars

 Flange root shape that is conformal with high rail
 Wheel/rail wear, rolling contact fatigue, rolling resistance

Slide 3 of 12
Wheel Profile: Design and Maintenance

• AAR-2A wheel profile for

freight cars
 Currently: alternate standard
for new and turned wheels
 Expect to require AAR-2A
for new and turned wheels
Jan 1, 2020
 Expect to obsolete AAR-1B
Jan 1, 2021

Slide 4 of 12
Wheel Profile: Design and Maintenance

• AAR-2A wheel profile for locomotives

 NUCARS® modeling of different wheel profiles on locomotives
 Curve/tangent, new/worn rail, wheel/rail friction, radial/non-radial trucks,
traction and braking forces

New/turned locomotive wheel

Worn locomotive wheel profiles:
profiles: Most very similar to AAR-1B
Conformal with worn high rail
Not conformal with worn high rail

Slide 5 of 12
 Wheel Profile: Design and Maintenance

• AAR-2A wheel
profile for
locomotives
 Considerations:
- Wheel truing cutter
heads
- More frequent truing
- Fuel savings
from reduced
rolling resistance

Slide 6 of 12
Wheel Profile: Design and Maintenance

• Optimal wheel rim thickness

 Thicker wheel rims for freight cars
- More opportunity to turn wheels
- Turned wheelsets less expensive
- Higher initial cost
- Heavier

Slide 7 of 12
Wheel Profile: Design and Maintenance
Potential
• Optimal wheel rim thickness Optimal
 Preliminary metric: percent
turned wheels applied
 Actuals from repair data
 Best case analysis based on
wheel removal causes and
associated wheel lathe cut depths
 Projected values scaled
to match actuals
 Consider economics

Slide 8 of 12
Wheel Profile: Design and Maintenance

• Hollow wheel wear limits

 Curving, hunting, rolling resistance
 Limit based on 1999 economic model
- NUCARS® simulations
- Limited testing at TTC
 Refresh economic model
- Update costs and benefits
- Incorporate large historical wayside data set
• Wheel hollow: Wheel Profile Detector
• Performance data: Truck Performance Detector (TPD) and Hunting Detector
- Relationship between tread hollow and lateral forces in curves is
complicated

Slide 9 of 12
Wheel Profile: Design and Maintenance

• Hollow wheel wear limits

 NUCARS® simulations to help guide analysis of TPD data
Hollow wheel on high rail of a curve can generate large lateral forces
4 mm
Hollow Hollow Mate
Wheel High Rail Wheel Wheel Low Rail
Flanging

Severe 2-Point Contact, Poor Steering

Mate wheel on asymmetrically worn wheelset can give low lateral forces
Mate 4 mm
Wheel Hollo Mate
Low Rail w Wheel High Rail
Flanging
Wheel
Conformal Contact,
Good Steering
Slide 10 of 12
Wheel Profile: Design and Maintenance

• AAR-2A wheel profile

 Implementation underway for freight cars
 Evaluating profile for locomotives
• Optimal wheel rim thickness
 Potential opportunities to improve on status quo for 36-inch wheels
• Hollow wheel wear limits
 Evaluating effects of hollow worn wheels
 Expansive data set using wayside detectors

Slide 11 of 12
Thank you!
Acknowledgements:
Railinc
TTCI:
Jack Schultz
Yuqing Zeng
RB Wiley
Chris Pinney
Tom Guins (retired)
Slide 12 of 12
Slide 12 of 12
Reconditioned
Bearing
Performance
Dustin Clasby
Steven Belport

©2019 TTCI
TTCI is a wholly owned subsidiary
Slideof1the AAR
of 14
Introduction to the Bearing Reconditioning Process

• Over 12 million bearings in revenue

Service
• Every bearing removed must go
through the reconditioning process
before it can be reapplied to an axle
 Removal
 Inspection
 Repair
 Reassembly

Slide 2 of 14
Introduction to the Bearing Reconditioning Process
BEARINGS APPLIED IN REVENUE SERVICE
Reconditioned New

New
9%

Reconditioned
91%

Slide 3 of 14
Introduction to the Bearing Reconditioning Process
REASONS FOR BEARING REMOVALS
Not Related to Bearing (Why Made (WM) -11) Wayside Alarm Other Bearing Specific Reasons

Wayside Alarm
Not Related to Bearing 3%
(Why Made (WM) -11)
90%
Other Bearing Specific
Reasons
7%

Slide 4 of 14
Introduction to the Bearing Reconditioning Process

REASONS FOR BEARING REMOVAL –

ONLY BEARING RELATED
Wayside Alarm Loose Bearing Components Derailment Govt Regulatory Seal Other

Other
9%
Seal
5% Wayside
Alarm
28%
Govt Regulatory
15%

Loose Bearing
Derailment Components
21% 22%

Slide 5 of 14
Inspection – Breakdown of bearing

Seals
Cone Assembly
Spacer

Cup

Seal Wear Rings

Slide 6 of 14
Inspection

• Visual
• Feeler Gauge

Slide 7 of 14
Reconditioning

Before Reconditioning

After
Reconditioning

Slide 8 of 14
Testing

• Twelve bearings selected with

similar defects (cup spalls)
• Performance was measured
with defect (before
reconditioning)
• Performance is currently being
measured after reconditioning
• Bearings will be run until failure

Slide 9 of 14
Testing - Comparison of Before and After Reconditioning

Before
Reconditioning

After Reconditioning
Slide 10 of 14
Results - Two Reconditioned Bearings

Slide 11 of 14
Results: First Failure

60,000+ miles
93,000+ miles
Spall formed

Slide 12 of 14
Summary

• Reconditioning Process
 Removal
 Inspection
 Repair
 Reassembly
• Performance based testing
of reconditioned bearings
 Temperature
 Vibration
 Life expectancy

Slide 13 of 14
 Thank You
We would like to acknowledge:
Constantine Tarawneh and his dedicated team
at the University of Texas – Rio Grande Valley
Steven Belport, TTCI

Slide 14 of 14
Slide 14 of 14