Tank car research

From: Transport Canada

This page contains abstracts of research of tank cars done by the Transportation of Dangerous Goods Directorate.

On this page

Abstract - Structural performance of TC-117 tank cars under derailment conditions – March 2, 2023

Transport Canada (TC) implemented the new TC-117 tank car design standard in 2015. This standard aims to make tank cars stronger and less likely to spill their flammable contents in a derailment. The TC-117 tank car can be built new (called TC-117J) or by updating (or retrofitting) an older tank car to add new required design elements (called TC-117R).

Transport Canada, in collaboration with the United States (U.S.) Department of Transportation (DOT) Federal Railroad Administration (FRA), has been working to quantify the safety benefits of the new standard. A TC study in 2018 assessed the risk of top fittings failure in derailments. In this current study, TC worked with Sharma & Associates, who used computer modelling to see what happens when TC-117J and various TC-117R tank cars derail at different temperatures and speeds.  Specifically, we looked at how many tank cars would be punctured and how many of the top fittings would be damaged.

Each computer simulation tested a unit train of 100 tank cars of one type, either TC-117J or TC-117R. We ran many tests at speeds between 5 and 60 mph (8 and 97 km/h). As steel becomes more brittle and easier to puncture at lower temperatures, we also developed a new way to predict the steel strength at lower temperatures. We estimated the number of punctures for all tank car designs and speeds at 20, -15, -25, and -40 °C.

We also ran simulations of a train where each tank car alternated with a rigid box car. Three (3) tank car designs were assessed: one (1) TC-117J and two (2) TC-117Rs. We wanted to see if tank cars hitting more sharp edges, in a more severe derailment, would change the number of punctures.

We found that:

  • at faster speeds, more cars derailed, more top fittings failed, and there were more predicted punctures
  • all TC-117 tank car designs had fewer predicted punctures than older TC-111 tank car designs, which had been studied earlier in a similar way by FRA and Sharma & Associates.
  • the new-built TC-117J tank cars had fewer predicted punctures than any of the retrofit TC-117R tank cars.
  • punctures increased between 5-10% when the temperature dropped from 20 to -40°C, in both the unit and mixed goods train scenarios.
  • the difference in puncture resistance between the best and worst performing tank car designs decreased in more severe derailments, but the TC-117J still provided a large improvement in performance over both the retrofit TC-117Rs and the legacy TC-111 tank cars.

This research highlights the safety benefits gained with the TC-117 design specification and the Rules Respecting Key Trains and Key Routes implemented by Transport Canada.

Learn more

Read a summary of the report: Structural performance of TC-117 tank cars under derailment conditions

To get a copy of the reports, please contact us.

TP: TP 15544E

ISBN: 978-0-660-46123-6

Catalogue: T44-3/32-2022E-PDF

Abstract - Validation of marshalling requirements for dangerous goods cars in a train: phases 1 and 2 – November 2, 2022

Canadian law has rules for where rail cars that are carrying certain dangerous goods can be placed in a train (called marshalling requirements). This helps because:

  • it keeps dangerous goods away from the train crew, and
  • it can separate goods that could spill and mix if the train derails

Transport Canada wanted to confirm whether Canada’s marshalling (i.e. car placement) rules offer a similar level of safety compared to the placement rules from other countries.

We completed this work over two (2) years and received two (2) stand-alone reports from our research partner, the National Research Council.

For 2019/2020, our goal was to find existing research on rail car marshalling and train make-up with dangerous goods. We also reviewed the dangerous goods placement rules from other jurisdictions, including:

  • United States of America
  • United Kingdom
  • European Union
  • Australia

We located 55 research reports to be reviewed, and found that most of these regions had similar regulations, with a few differences like:

  • United States
    • For trains made up of a single dangerous good (unit train), they are required to have at least one buffer car between locomotives and dangerous good cars while Canada does not require any buffer car (as of the date of this publication)
    • Since 2002, there is a different number of required buffer cars for mixed freight trains (5 U.S. to 1 Canada)
  • Australia
    • Food and food packaging are required to be separated from class 8 dangerous goods
  • Europe/UK
    • More detailed table for separation requirements in Europe/U.K. regulations for the marshalling of dangerous goods

In the second year of work (2020/21), we did an in-depth literature review of the 55 research reports and focused on how the placement of dangerous good cars and use of buffer cars or separation requirements within a train (marshalling) affected the risk of derailment and severity of accidents.

In addition to the literature review, the National Research Council used Canadian and U.S. incident reports to compare derailments that involved multiple cars to see how the results differed based on the differences between the U.S. and Canadian buffer-car requirements. To compare the derailments, data was collected on the location of the first car to derail, the number of cars to derail and the number of accidents per year.

The literature review did not find a reason for the current placement and buffer car requirements and was not able to find an ideal distance between dangerous good rail cars and locomotives or crew. There wasn’t a major difference in the outcomes of derailments before the Canadian buffer car requirements changed in 2002, or from 2002 to 2020.

We did find evidence that reducing the number of dangerous good cars in the first portion (such as cars one (1) to five (5)) of the train may:

  • lessen the risk of a dangerous good car derailing in a front-of-train derailment
  • lower the number of dangerous goods cars that could spill in an accident
  • lessen the risk to the crew in an accident (because there would be more space between the crew and dangerous good cars)

However, there was not enough information to quantify the impact of this potential change.

Learn more

To get a copy of the reports, please contact us.

2019/20 Report
Title: Dangerous goods tank car marshalling analysis: phase 1 literature search, jurisdictional review
TP: TP 15492E
ISBN: T44-3T44-3/25-2021E-PDF
Catalogue: 978-0-660-40380-9

2020/21 Report
Title: Validation of marshalling requirements for dangerous goods cars in a train: phase 2
TP: TP 15472
ISBN: 978-0-660-38587-7
Catalogue: T44-3/22-2021E-PDF

Abstract - Finite element analysis of rail tank car hard coupling - July 14, 2022

Linking two (2) rail cars together is called “coupling”. When coupling happens at too high a speed, this is called “hard coupling”. The Transportation of Dangerous Goods Regulations has speed limits and other requirements for coupling rail tank cars carrying dangerous goods.

Transport Canada wanted to study whether these speed limits help prevent damage during hard coupling. This work was done with Natural Resource Canada’s CanmetMATERIALS research centre.

First, we reviewed existing regulations, research, and incidents related to hard coupling in Canada. We did this to find key areas that could benefit from more research. The three (3) main areas of research we identified were:

  • updating a finite element impact model with details of rail tank car stub sills, adding material modelling to include low temperature effects, and assessing the importance of A572 steel for stub sill modelling
  • assessing the effects of rail tank car draft gear performance under a wide range of conditions: age, type, operating temperature, and impact velocity (speed)
  • assessing how steel fatigue and embedded flaws affect hard coupling failures, and how effective inspections (type of inspection and how frequent) are at detecting small cracks before they fail

This current study focuses on the first area. It used experimental data and finite element computer modelling to simulate a tank car coupling event with updated impact and material models.

Material modelling was based on existing test data of TC128B at various temperatures and strain-rates from CanmetMATERIALS. We did limited mechanical tests on A572-50, a common stub sill material, and compared the results to TC128B. We found that the existing models for TC128B are adequate for the purposes of this study.

We used some hard coupling test results from the United States Department of Transportation’s Federal Railroad Administration to calibrate and validate the model. We also modelled the worst-case scenario tested by the United States: full tank cars coupled at 10 mph with steel friction draft gears.

The models showed no sign that a single hard coupling event at speeds of 6 to 10 mph would result in any noticeable damage to an undamaged tank car at the temperatures studied (25°C and -40°C).

While the results of this study suggest current coupling speed limits are effective, the model used was quite simplified.

Possible areas of future research:

  • modelling higher impact speeds and lower temperatures to see when single impacts could lead to damage
  • investigating how embedded flaws in the steel, along with low temperatures, vertical coupling force, and repeated coupling could make damage more likely
  • using a damage tolerance analysis to determine the size of a flaw or crack that will lead to failure (in other words: the critical crack size), and compare this to cracks caught in common inspection methods
  • learning more about how damage occurs and spreads in TC128B weld material

Learn more

Read a summary of the report: Finite element analysis of rail tank car hard coupling

To get a copy of the report, please contact us.

TP: TP 15512E
ISBN: 978-0-660-41612-0
Catalogue: T44-3/27-2022E-PDF

Abstract - Evaluation of current tank car TC128B steel weld performance - July 14, 2022

Since 2015, Transport Canada and Natural Resources Canada’s CanmetMATERIALS (CMAT) have been investigating the types of steel often used in rail tank cars.

We started this project in 2019 to complement the work we’d already done on Strength, Creep, and Toughness of Two Tank Car Steels. This project improved our understanding of how the steels used in rail tank cars perform in incidents at high and low temperatures.

We tested the weld joints between sections of a non-pressure rail tank car (DOT 117 specification). We did experiments to determine:

  • composition
  • strength (micro-hardness, tensile strength, and toughness)
  • weld performance

The steels’ composition and strength were tested at room temperature (23°C). Welds were evaluated through two (2) different tests:

  • tensile properties testing (done between 850°C and -60°C)
  • charpy impact testing (done between 25°C and -80°C)

These tests were chosen since they’re often used in metallurgy (the science of metals) to characterize steels.

Charpy impact testing is listed in the Association of American Railroads, Specifications for Tank Cars M-1002 (2014). This specification requires Charpy tests at 34°C and -46°C for certain types of rail tank cars. We chose the testing temperatures so we would have enough data points to create a profile of the steel in a wide range of temperatures.

This study showed that the sample steel met the composition requirements for TC128B tank car steel. We also found that the hardness and welds performed as expected. The tensile strength of the samples increased as the temperature decreased. The toughness of the samples, measured by impact testing, decreased as the temperature decreased, and the toughness of the weld samples was lower than that of the base material. The sample steel was deemed to be compliant with the standards for a general service non-pressure rail tank car (noting that if the weld sample had come from a pressure tank car or a special commodity non-pressure tank car, the Charpy test would have failed).

Learn more

Read a summary of the report: Evaluation of Current Tank Car TC128B Steel Weld Performance

Full reports:

TP: TP 15515E; TP 15470E
ISBN: 978-0-660-42411-8; 978-0-660-38585-3
Catalogue: T86-74/2022E-PDF; T44-3/20-2021E-PDF

Abstract – Assessment of Alternating Current Field Measurement Non-Destructive Testing (ACFM NDT) for use on tank cars – June 7, 2022

Non-destructive testing techniques are often used to find defects during routine tank car inspections. Alternating current field measurement is a non-destructive testing technique that tank car manufacturers want to use to inspect butt and fillet welds. Alternating current field measurement tests have many advantages:

  • they are easier for inspectors to set-up
  • can be used remotely, and
  • are easy to clean-up, which makes inspections faster

In 2020, Transport Canada (TC) and our research partner Natural Resources Canada (NRCan) prepared an internal report as an initial assessment of potential non-destructive testing techniques for tank car inspection. Based on the findings in that report, TC partnered with the National Research Council Canada (NRC) in 2020-21 to evaluate the possibility of standardizing the use of alternating current field measurement test to inspect tank cars. For this study, the alternating current field measurement test method was evaluated and compared against two (2) other methods: magnetic particle and liquid penetrant testing.

To complete this research, we spoke with experts in rail tank car inspections, materials, and non-destructive testing experts from NRCan CanmetMATERIALS (CMAT), the NRC and the U.S. Federal Railroad Administration.

The U.S. Federal Railroad Administration provided us with sample rail tank car plates from their defect library that we used to evaluate and compare different testing techniques. As well, four (4) certified and well-trained inspectors examined each sample plate. These inspectors had no prior knowledge of the location, number, or size of flaws present in each plate.

We used procedures for liquid penetrant and magnetic particle tests based on standard industry practices. Non-destructive testing experts developed a test procedure for the use of alternating current field measurement tests on the sample plates.

For every test, inspectors used the correct test procedures and recorded the location and length of any flaws they found during inspection. For alternating current field measurement tests, inspectors also determined the depth of the flaw.

Research results

We compared the inspection results to evaluate the performance of each testing technique. The indication count of a defect and the hit ratio by indication length were used to evaluate the techniques.

We found that all testing methods were able to detect surface-open cracks that were included on the sample plates, and all methods could estimate defect length. Each testing method has its own advantages and disadvantages since each method is based on different physics and has different sensitivity for different types of defects

This study found that alternating current field measurement (ACFM) tests can successfully detect a wide range of surface-open crack lengths and depths compared to the usual inspection techniques. We also found that the level of an inspector's experience was a major factor in the accuracy of inspections.

Learn more

Read a summary of the report: Assessment of Alternating Current Field Measurement Non-Destructive Testing (ACFM NDT) for use on tank cars 

To get a copy of the report, please contact us.

TP: TP 15513E
ISBN: 978-0-660-41613-7
Catalogue: T44-3/28-2022E-PDF

Abstract – Tank Car Fire Failure Assessment using Combined Models – April 26, 2022

In 2018, Transport Canada and Natural Resources Canada's CanmetMATERIALS team started work on modelling a rail tank car in high-temperature fires. We used a computer program called “Abaqus” to find out:

  • how long a rail tank car can survive in a fire
  • the location where the steel fails

This work used material properties from our previous work on common rail tank car steels as listed in Strength, Creep, and Toughness of Two Tank Car Steels (2019). That project improved our understanding of how the steels used in rail tanks car perform in incidents at low and high temperatures.

To understand how different high temperature cases affect the tank car model, we used finite element computer models. The first case we created is called the “base case” which used a typical TC-117 rail tank car loaded with a light crude oil. Data on the heat transferred from a fire to the tank car was taken from thermodynamic modelling work done by Natural Resources Canada's CanmetENERGY team.

A total of 34 cases were tested, with the following variables (anything that can change or be changed) switched, one at a time:

  • whether the tank car was level or had rolled over
  • the fire's temperature
  • how full the tank car was
  • the set pressure of the pressure relief valve (PRV)
  • whether the PRV was blocked or not
  • the thickness of the shell steel
  • the type of crude oil in the tank car (lading)
  • the emissivity of the fire and steel
  • whether or not the tank car had thermal protection

Another part of this work was to develop a simplified engineering model using Microsoft Excel and compare the results to the finite element model. The engineering model was updated to use peak tank wall temperature and internal pressure over time as inputs, and calculate plastic strain, creep strain, and failure time.

We also compared the finite element model results to the Analysis of Fire Effects on Tank Cars (AFFTAC) model used by the North American rail tank car industry. We defined a light crude oil lading in AFFTAC and used the tank car design from the finite element model to see if they had the same result.

Research results

The CanmetMATERIALS engineering model and finite element model had similar creep strain and failure time for all cases.

When we compared the results from all three (3) models:

  • all cases saw rail tank cars survive at least 100 minutes in a fire (as required by Transport Canada's TP 14877)
  • 32 out of 34 cases had the same results until the end of the 712-minute simulation time, and the tank cars did not fail
  • Two (2) extreme cases (fully-blocked pressure relief valve and no thermal protection) caused the tank cars to fail within the 712-minute simulation time in the CanmetMATERIALS models, but not the AFFTAC model

Transport Canada's report includes details and discussion on the remaining differences between the three (3) models, but no major concerns were found.

Learn more

Read a summary of the report: Tank car fire failure assessment using combined models

Full report: Tank car fire failure assessment using combined models

TP: TP 15493E
ISBN: 978-0-660-40379-3
Catalogue: T44-3/24-2021E-PDF

Abstract – Rail tank cars exposed to fire: Literature review of crude oil, condensate and ethanol behaviour – August 8, 2018

NRC technical report A1-005795-01.1

In an effort to better understand how tank cars carrying crude oil behave in fires, Transport Canada asked the National Research Council of Canada (NRC) to review related open literature. The NRC’s literature review focused on the following topics:

  • pool fires
  • large obstructions and large heat absorbing objects in fires
  • behaviour of crude oil, condensate and ethanol fires
  • fire modeling
  • behaviour and modeling of complex mixtures such as crude oil.

The report describes general characteristics of pool fires including factors that influence the heat transfer to engulfed objects such as fuel type and wind. The report goes on to describe the characteristics of pool fires from a number of sources and some pool fire modeling work. The report found that it is very difficult to predict how complex mixtures containing many components (such as crude oil) change with temperature and pressure. To overcome this difficulty, a number of numerical methods based on experimental investigations have been developed.

The NRC also developed a research plan to address gaps in the available literature. This plan describes a series of testing from small flammable liquid pool fire tests to progressively larger tests of containers mimicking tank cars in flammable liquid pool fires. Investigation into how pressure relief devices meant to release gas behave when expelling viscous liquids are suggested. Fire modeling and crude oil behaviour modeling are also discussed as areas requiring further research.

Full report: Rail tank cars exposed to fire: Literature review of crude oil, condensate and ethanol behaviour

Please direct technical questions regarding the report to the NRC. If you would like other information regarding the full report, please contact us.

Abstract - Risk evaluation of tank car top fittings breach in derailments – August 8, 2018

Transport Canada asked Sharma & Associates, Inc. to investigate how well various tank car top fittings protection strategies performed in derailments. Sharma & Associates, Inc. used a combination of models they originally developed for the United States Department of Transportation Federal Railroad Administration to analyse the performance of individual tank cars and the forces involved in whole train derailments.

Top fittings are devices mounted on the top of a tank car such as:

  • fill gauges
  • discharge pipes
  • thermometer wells
  • safety vents
  • other connections or valves.

These fittings are strong but are typically not designed to survive a derailment. They are often covered by some form of protection. The strength of the protection depends on the risk posed by the material being carried in the tank car. Tank cars carrying products such as corn syrup or clay slurry will have fairly thin covers that only protect against vandalism or the weather. Tank cars carrying higher risk materials, such as some flammable liquids, use stronger covers to help protect against damage to top fittings in derailments. Tank cars carrying very high risk products, such as toxic gases, use the strongest covers which are designed to help prevent damage to top fittings in derailments.

Some of top fittings protection strategies investigated in this report include:

  • stronger covers
  • reducing train speeds
  • different train braking systems
  • shorter trains.

For each of these cases, tank car performance models and derailment simulations were combined to estimate how many top fittings would be damaged. The results of these simulations were compared against real accident data. Reasonably good correlation was found between the simulation results and the accident data. Overall, the simulations found that slower trains, stronger covers, and faster braking systems reduced the number of damaged top fittings.

If you would like more information regarding the full report, please contact us.

Abstract - Natural Resources Canada tank car steels literature reviews – August 8, 2018

At Transport Canada’s request, and with Transport Canada’s support, Natural Resources Canada’s (NRCan) Canmet MATERIALS group conducted reviews of existing literature on the properties of two common tank car steels. These tank car steels were TC128B and ASTM A516 Grade 70. TC128B is commonly used in the shells of tank cars that carry dangerous goods. ASTM A516 Grade 70 is used in the shells of general service tank cars though it is sometimes also found in the shells of tank cars that carry dangerous goods. The reviews were titled:

  • A Review of the Strength and Fracture Toughness Properties of Two Tank Car Steels: TC128B and A516-70
  • Mechanical Properties of Tank Car Steel at Flame Temperature and Modeling of Failure – A Review
  • Internal Corrosion of Rail Tank Cars

The literature reviews focused on the following steel properties, respectively:

  • the resistance of steel to cracking at various temperatures (fracture toughness)
  • how the strength of steel changes at elevated temperatures (including creep properties)
  • a primer on corrosion focusing on corrosion arising from transporting crude oil.

The fracture toughness report summarizes the results of several studies investigating fracture toughness in both TC128B and ASTM A516-70 steels using a variety of standards and test methods. The report also describes some newer techniques to test the fracture toughness of steels. A number of the reports reviewed were older. Some of the data presented in these reports may not represent the performance of current steels produced using more modern steel manufacturing practices and with the composition of the steels that are currently found in the standards.

The review of the performance of tank car steels at elevated temperatures summarized the few high temperature studies conducted on TC128B steel. No such studies were found for ASTM A516 Grade 70 steel. The report also reviews some common software models of tank car fire performance at elevated temperatures.

The corrosion report describes the basics of corrosion with a focus on likely mechanisms found in tank cars. Some circumstantial evidence of corrosion in some tank cars carrying crude oil was found but a more thorough investigation is required to determine its prevalence and causes.

If you would like more information regarding the full reports, please contact us.

Abstract – Strength, creep, and toughness of two Tank car steels - August 20, 2019

In accidents involving rail tank cars, these tanks may fail from situations like punctures or pool fire scenarios. Understanding the material properties of different steels can help us improve our understanding of how tank cars perform during accidents.

In 2015, Transport Canada partnered with Natural Resources Canada’s CanmetMATERIALS (CMAT) to review existing literature on tank car steels. CMAT was able to identify research gaps on material property data. With the support of CMAT, Transport Canada ran laboratory tests from 2016 to 2018 to collect data for two common tank car steels.

In this study, TC128B and ASTM Grade 70 (A516-70) tank car steels were tested for:

  • composition
  • microstructure
  • strength between -80°C to 800°C
  • creep: The loss of strength at temperatures above 500°C over a long period of time
  • toughness: The resistance of steel to cracking.

The study showed that the composition of TC128B, its strength, and its toughness meet the Association of American Railroads (AAR) specifications for tank car steels.

At room temperature the steel forms diagonal bands called Lüders bands, which is common for low-carbon steels. As the temperature of TC128B increases, the strength decreases and the steel creeps much faster. The materials strength has little dependence on the direction it was tested.

Overall, the material properties of A516-70 are similar to TC128B. Differences between the steels include A516-70 having lower strength, higher toughness, and larger particles in its microstructure than TC128B.

From this study, several equations relating the strength, creep, and toughness properties of TC128B and A516-70 steel were developed.

Full report: Strength, creep, and toughness of two tank car steels, TC128B and A516-70

ISBN: 978-0-660-32009-0

To obtain a copy of the report, please contact us.

Contact us

Safety Research and Analysis Branch
Transportation of Dangerous Goods Directorate
Transport Canada
Email: TC.TDGScientificResearch-RecherchescientifiqueTMD.TC@tc.gc.ca

Report a problem with this page

Date modified: