A Review of the Department of Transportation Plan for Analyzing and Testing Electronically Controlled Pneumatic Brakes (2017)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

20 tween modeling approaches might be explained, at least in part, by the assumptions made by the modelers. The DOT analysis and test plan, described next, is a proposed way to evaluate whether critical assumptions are justified. DOT ANALYSIS AND TEST PLAN In carrying out its task, the committee considered DOT’s Analysis and Test Plan to Assess the Effec- tiveness of ECP Brakes in Reducing the Risks Associated with High-Hazard Flammable Trains. The results of the analysis described in the plan are intended to update the consideration of ECP brakes in the RIA for the HHFT final rule, as required by Congress in the FAST Act. The testing is intended to improve the model- ing approach that DOT will use in the analysis. A summary of the plan is provided below. The entire plan is presented in Appendix E. Analysis The analysis portion of the plan addresses six aspects of calculating ECP brake effectiveness: • Assess the sensitivity of calculations of ECP brake benefits to various values of the coefficient of fric- tion between a sliding derailed tank car and the ground until the car comes to rest. • Calculate the puncture resistance of a DOT-117 tank car from various tank material properties, such as stiffness, and validate the model by using data from full-scale impact tests. • Calculate the benefits of ECP brakes on a curved track that is representative of the distribution of curves on the rail network. • Analyze alternative derailment initiation events (such as broken rails or broken wheels) and model the additional scenarios if they are found to be significantly different in nature from an applied lateral force, which was modeled for the HHFT RIA. • Compare the sequence of events (including BPP changes throughout the train, emergency brake appli- cations, and locomotive engineer responses) in a derailment of a train fitted with an EOT device with that of a train outfitted with DP at the front and rear. • Validate the modeling approach used for the HHFT RIA by modeling previous real-world derailments and comparing the results with what actually happened, such as the number of punctures and distances traveled by cars from derailment to rest. Full-Scale Testing The plan includes two types of full-scale physical testing to obtain measurement data for its analysis portion. One test determines the coefficient of friction between a sliding tank car and the ground under a va- riety of conditions, including wet, dry, and frozen ground. The other is a test of the puncture resistance of DOT-117 tank cars.50 IDENTIFYING KEY CONSIDERATIONS FOR DOT PLAN Although the parameters and other factors chosen by DOT for analysis and testing are relevant to con- sideration of the effectiveness of ECP brakes, DOT’s plan provides no justification as to why those items are more important than others that could have been considered. This section discusses the development of a merged database and the use of multivariate regression analysis to inform the identification of factors that are worthy of testing or of additional measurements. Such a database and analysis could strengthen DOT’s ap- proach for considering ECP brakes in the HHFT final rule. 50DOT carried out the puncture resistance test at the Transportation Technology Center in Pueblo, Colorado, on September 28, 2016. Committee members observed the conduct of the test and will consider the results in the Phase 2 report.

21 Analysis of Existing Accident Data In identifying modeling factors to be analyzed or tested, it would be prudent and efficient to extract all the reasonably obtainable information from the existing data on train derailment accidents that have already produced tens of thousands of damaged tank cars. Logically, until all that is known about tank car derailment accidents that produce spillage is identified, efficient construction of a test program for filling in the knowledge gaps about such accidents is impossible. Therefore, careful analytical effort is required in exploit- ing the full value of the data for model and simulation purposes. Perhaps even more important, the relative benefits to be gained by filling in different information gaps through physical testing are likely to differ with respect to improving the predictive power of a model. The costs of the physical testing that would be required to fill in the various gaps would often be widely dispar- ate. While almost any improvement in the accuracy of a simulation or model component will improve the accuracy of the subsequent predictions, limited testing resources dictate that those resources be used for re- fining the accuracy of the most powerful predictive parameters first. As indicated in the HHFT final rule, PHMSA estimates that over a 20-year period, an average of 15 in- cidents per year involving the derailment and release of crude oil or ethanol will occur. In some cases, knowledge that could help improve models or simulations could be obtained through enhanced accident in- vestigation for less cost than obtaining such knowledge through full-scale physical testing. FRA’s RAIRS database,51 which spans decades, contains information on thousands of tank car derail- ment accidents. The database includes information on the spillage of crude oil or ethanol, among other data, and represents an invaluable historical resource. FRA performs ongoing analysis of the RAIRS database and has used the results for past assessments of safety-critical train control systems.52 When the RAIRS tank car derailment accident records did not provide sufficient detail, particularly about tank cars and the nature of their damage, FRA supplemented the records with tank car–specific data from another database, in one case the Universal Machine Language Equipment Register.53 The committee judges that significant benefits in addressing the technical issues related to the latest de- railment model and their implications for ECP brakes would be realized by supplementing the RAIRS tank car derailment accident records with tank car–specific information, particularly on damage. Such information is available in the tank car accident data of the Railway Supply Institute (RSI) and AAR. A merged database of tank car derailment accidents could be created. The RSI-AAR database of damaged tank cars has been compiled by the RSI-AAR Tank Car Safety Research and Test Project.54 The data include information on damage to tank car–specific components sustained through train accidents and whether there was loss of car contents. Such a merged database might have many uses beyond the current project. Using Accident Data to Build a Multivariate Regression Model A multivariate regression model developed from the merged database would be useful for examining how the response variable (tank car spillage) relates to a set of predictor variables (specific circumstances of the accidents). Identification of the most appropriate regression form (such as multivariate linear regression 51http://safetydata.fra.dot.gov/officeofsafety/default.aspx. 52Mokkapati, C., T. Tse, and A. Rao. A Practical Risk Assessment Methodology for Safety-Critical Train Control Systems. DOT/FRA/RDV-09-01. Office of Research and Development, FRA, U.S. Department of Transportation, July 2009. 53Hazel, M. E., Jr. Tank Car Accident Data Analysis. DOT-VNTSC-FRA-91-3. Office of Research and Develop- ment, FRA, U.S. Department of Transportation, June 1991. 54Treichel, T. T., J. P. Hughes, C. P. L. Barkan, R. D. Sims, E. A. Philips, and M. R. Saat. Safety Performance of Tank Cars in Accidents: Probability of Lading Loss. RSI-AAR Railroad Tank Car Safety Research and Test Project, AAR, Washington, D.C., 2006.

22 or multivariate nonlinear regression) is beyond the scope of this discussion but would become apparent in the course of the analysis. Various publications are available to inform this kind of data analysis.55 Such a model would illustrate how well the likelihood of the spillage outcome can be predicted by as- signing appropriate weight to the various factors that might be involved in any particular derailment—for example, initial train speed, derailment position, terrain, track grade, brake type, and car NBR. Insights into the factors exerting the most powerful influence on the outcome could be provided. Such analysis often indi- cates which factors can be virtually ignored and which are crucial in predicting the outcome of interest. It can also provide insight into which predictors are strong but known with insufficient accuracy and therefore are worthy of testing or additional measurement. Such modeling of accident data is not new to DOT. Multivariate analysis of main-track train collisions was used by FRA’s Collision Analysis Working Group to analyze the injury risk of employees on the basis of their decision to remain with or exit a locomotive in the event of a collision. “This multivariate technique controls for confounding variables while testing the effect of interest—whether the employee’s decisions to exit or stay, given collision certainty, changed the risk of injury or fatality.”56 Numerous hypotheses could be tested by a rigorous multivariate analysis of the existing accident data. In view of the historical use of EOT braking systems, the obvious starting point would be whether and to what degree the presence or absence of such a system on the train affected the likelihood of spillage in the subsequent derailment. Multivariate multiple regression analysis, particularly with so many data points, would be able to control for other variables that are expected to have a strong impact on this outcome, such as tank car type, train speed, and train length. Alternatively, factors could be considered in concert to explore their relative influence. The multivariate analysis could immediately and independently confirm whether the presence of an EOT braking system could cause an approximately 20 percent reduction in the likelihood of spillage, relative to pneumatic brakes, as predicted by the DOT–Sharma modeling. Once the model accurately predicts the outcomes that are known, it could be modified to address the ECP braking question with more confidence. For example, this approach could be used to indicate the likely value of testing the coefficient of friction measured or assumed in various accidents for greater precision. DOT’s analysis and test plan includes a test that involves sliding a tank car across soil to measure a coefficient of friction. If the tank car sliding coefficient of friction has been demonstrated to have a powerful influence on the predicted outcome in an analysis of real-world tank car derailment accidents, that test might be a prudent use of testing resources. However, to the committee’s knowledge, this has not been demonstrat- ed. Furthermore, whether the sliding coefficient of friction over soil is more or less important than the soil’s mechanical resistance to tillage is not clear from first principles. Many features on a tank car actually plow 55Barkan, C. P. L. Improving the Design of Higher-Capacity Railway Tank Cars for Hazardous Materials Transport: Optimizing the Trade-Off Between Weight and Safety. Journal of Hazardous Materials, Vol. 160, No. 1, 2008, pp. 122–134. Saat, M. R., and C. P. L. Barkan. Generalized Railway Tank Car Safety Design Optimization for Hazardous Materi- als Transport: Addressing the Trade-Off Between Transportation Efficiency and Safety. Journal of Hazardous Materi- als, Vol. 189, Nos. 1–2, 2011, pp. 62–68. Liu, X., M. R. Saat, and C. P. L. Barkan. Analysis of Causes of Major Train Derailment and Their Effect on Acci- dent Rates. Transportation Research Record: Journal of the Transportation Research Board, No. 2289, 2012, pp. 154– 163. Liu, X., M. R. Saat, X. Qin, and C. P. L. Barkan. Analysis of U.S. Freight-Train Derailment Severity Using Zero- Truncated Negative Binomial Regression and Quantile Regression. Accident Analysis and Prevention, Vol. 59, 2013, pp. 87–93. Liu, X., M. R. Saat, and C. P. L. Barkan. Probability Analysis of Multiple-Tank-Car Release Incidents in Railway Hazardous Materials Transportation. Journal of Hazardous Materials, Vol. 276, 2014, pp. 442–451. Liu, X., and Y. Hong. Analysis of Railroad Tank Car Releases Using a Generalized Binomial Model. Accident Anal- ysis and Prevention, Vol. 84, 2015, pp. 20–26. 5665 Main-Track Train Collisions, 1997 through 2002: Review, Analysis, Findings, and Recommendations. Collision Analysis Working Group Final Report. FRA, Aug. 2006, page xx.

23 the soil as well as the track and ballast (usually crushed granite) as opposed to merely sliding over it. There- fore, improvement of the ability of the DOT–Sharma approach to estimate the retarding force resulting from a derailed tank car plowing into the ground might be more important than a focus on the coefficient of fric- tion. The locations of thousands of past tank car derailment accidents are known. The soil conditions and character in some may have changed since the date of the accident, but they certainly have not in many oth- ers. Data on soil character, surface sliding coefficient, and tillage resistance could likely be collected from a number of past accident sites at a cost substantially less than one tank car sliding test. A reduced set of sup- plemented accident records could then be subjected to a multivariate analysis to observe the strength of soil character in predicting the outcome of the accident. Investigation of future accidents could be augmented to capture this variable at additional modest cost. USING ACCIDENT DATA FOR MODEL VALIDATION DOT’s analysis and testing plan includes model validation (see Appendix E, Item 2.6). In a preliminary effort to validate the DOT–Sharma modeling and simulation approach, DOT qualitatively compared the out- put of 18 simulations of tank car derailment accidents run at three initial train speeds with three images of actual tank car derailment pileups. The simulations’ output of the expected number of cars derailed was por- trayed graphically against a background of data points from the RAIRS database (1992–2013) that had been filtered for scenarios similar to the simulations.57 As mentioned above, the validation effort involved only derailments within a narrow domain of the simulations that were obtained from the database.58 However, as indicated in DOT’s analysis and testing plan, comparison of the model with many known derailment out- comes is required for increasing confidence in the analysis and the conclusions that are reached from the simulations. The validation process needs to involve comparison of modeling results of the outcome of an individual derailment event with the actual outcome of that event. Such a comparison should be conducted with a rea- sonably large number of real-world accidents. It is important for adequate physics-based metrics to be the basis of comparisons between the event and the model predictions. Examples of physics-based metrics are the number of tank cars derailed and the number of tank cars punctured. The metrics used will depend on the accident data available from the merged database. If the merged database suggested earlier is created, the number of punctures that actually occurred in derailment accidents could be readily compared with the model predictions of that parameter under the same speed and track conditions. The appropriate number of validation exercises cannot be known in advance. A good starting point would be to compare the model with known derailments over a range of initial speeds common in freight train operation—for example, from 10 to 50 mph, in 5- to 10-mph increments—with different train lengths and in the presence of other common operating conditions. The conditions of each derailment would have to be similar to those that the model can approximate, in terms of both train consist and track infrastructure, such as the terrain, ties, and ballast. The modeling parameters would have to be selected in accordance with a thorough analysis of the accident data available. Because relatively few ECP-equipped unit trains are in use in the United States, little derailment infor- mation is available on them. (To date, only coal hoppers and intermodal cars have been equipped with ECP brakes and operated in unit trains.) However, thousands of derailments of trains with pneumatic brakes and conventional tank cars have been investigated, as have hundreds (if not thousands) of derailments of trains with an EOT braking system also composed of conventional tank cars. The appropriate statistical tests would have to be done on these two populations to ensure that they are similar. The remainder of this discussion assumes that their sheer size might ensure their similarity. Even if they are not similar, the following discus- 57Letter Report: Objective Evaluation of Risk Reduction from Tank Car Design & Operations Improvements— Extended Study. Sharma & Associates, March 2015. 58Final RIA for the final HHFT rule, page 72.

24 sion still might be relevant if, for example, there is some structural bias, such as the trains using EOT brakes having a higher probability of being formed of better tank cars. If these two large groups of accidents are similar populations, the DOT modeling prediction of EOT braking system benefits over pneumatic braking without EOT devices could be compared directly with the observed benefits, if any, of EOT braking in real tank car derailment accidents involving both types of brak- ing in the merged database. Close agreement would provide significant validation of the DOT–Sharma mod- eling prediction for this important metric and might lead to much more confidence in the ECP benefit projec- tions. Of course, a significant difference would indicate that the modeling needs more work. Alternatively, the DOT–Sharma approach could be applied to any combination of tank car type, train speed, and braking for which robust accident data exist, so that its predictive accuracy could be measured against what actually occurred. Even if the available accident data are incomplete or lack detail, they can still constitute a powerful tool. First, a simple crash scenario for which the outcome is known and has been meas- ured is simulated—for example, the measured crush when the vehicle crashes straight into a rigid steel barri- er. After the model has been tuned or upgraded as necessary to predict or simulate a simple scenario accu- rately, more complicated scenarios are simulated. Each time, the simulation of a known derailment with measured outcomes is used to evaluate how accurately the model predicts these important outcomes. Evidence that the model at least predicts some simple tank car train derailments with known outcomes with acceptable accuracy would increase confidence in use of the model to predict the outcome of the same accidents had ECP brakes been present. FINDINGS AND RECOMMENDATIONS ECP Brake Force Propagation in Emergency Applications 1.1. Finding The DOT–Sharma modeling approach includes no delay in the emergency application of ECP brakes. All train cars are assumed to initiate emergency braking at the beginning of the simulation; a latency time to detect loss of BPP and for signals to be received at all cars to initiate emergency braking was not included in the modeling. The DOT–Sharma modeling approach assumed that the ECP brake force increases linearly from zero to maximum value in 9.6 seconds. 1.2. Finding The DOT–Sharma modeling approach interprets AAR Standard S-4200 (Paragraph 4.4.4.1, Multiple CCDs, Trailing HEUs, or EOT Has Critical Loss) as not allowing a delay for initiating the emergency application of ECP brakes. Some test rack data (see Appendix F) exhibited a delay between the time of the loss of BPP and the beginning of the rise in BCP. 1.3. Finding The results of train car test rack simulations provided to the committee at its request represented the re- sponses of both an ECP-only system and an ECP-OL system. 1.4. Finding While there was variability in the test rack results, some results suggest that during the first few sec- onds of emergency brake application in response to an air hose separation, initial cars approaching the POD in a train with ECP brakes might experience less braking force than would cars in a train with pneumatic brakes. Some test rack results exhibited a lag during the first 1.5 seconds for ECP brake propagation followed by a rise to full pressure in the BC in 9 to 10 seconds. Although the committee does not consider the test rack results it received to be definitive with respect to emergency brake per- formance, the results point to key aspects in need of further testing.

25 1.1. Recommendation To ensure that appropriate pressure–time relationships are included in comparative analyses of brak- ing systems, the performance of ECP systems relative to DP and EOT systems should be simulated by using one or more AAR-certified test racks. The derailment-initiating event for ECP testing should comprise a synchronized separation of the air hoses and car-to-car electrical connectors. Tests should be conducted at multiple car positions throughout the test racks. Tests should use CCDs of both the ECP-only and the ECP-OL type. The tests should be designed to characterize and contrast the amount of braking power that ECP brakes and pneumatic brakes augmented with EOT devices or a DP config- uration develop from the initial train separation to the point at which all of the brakes on the train are fully applied. Some derailments are a result of hazardous conditions in front of a train (such as obstructions, misa- ligned switches, and broken rails). However, analysis of an emergency brake application that is initiated by a locomotive engineer at the front end of the train is not pertinent to the effectiveness of brake con- trol systems that are initiated after a first car, or one of its wheels, derails. 1.2. Recommendation DOT should measure the time required for the emergency-mode application of ECP brakes and pneu- matic brakes augmented with EOT devices or a DP configuration from initial train separation to the point at which all of the brakes on the train are fully applied by testing full-scale locomotive and reve- nue cars that are in service. The test should be conducted on standing trains and ones that are moving slowly. Both ECP-only brakes and ECP-OL brakes should be examined. Emergency braking conditions should be initiated by simultaneously disrupting the BP and electric line connections. EOT and DP ra- dio-system operational delay times should be collected from field studies and then applied in test rack simulations. DOT Analysis and Test Plan 2.1. Finding DOT’s analysis and test plan provides no justification as to why the factors chosen for consideration are the most important in predicting brake performance. 2.2. Finding Data on train derailment accidents that have resulted in thousands of damaged tank cars are available for analysis to guide the plans for modeling and large-scale physical testing. 2.1. Recommendation Before implementing the plan, DOT should use multivariate analysis to inform its development and to ensure that it is focused on the factors that have significant impact on model predictions. 2.2. Recommendation FRA’s RAIRS tank car derailment accident records with tank car–specific information (particularly concerning damage) as found in the RSI-AAR tank car accident database should be used to create a merged database for tank car derailment accidents.

26 2.3. Recommendation A multivariate regression model should be developed from the merged database to examine how tank car spillage relates to a set of specific circumstances of the accidents [e.g., number of cars derailed, number of broken couplers, the nature of damage to the track structure (including the length of dam- aged track), the speed of the trailing locomotive at the start and during the accident (if available), and the magnitude and location of impact forces for each car in relation to the recorded postaccident dam- age]. The results should be examined for insights into which analysis predictors are strong but are known with insufficient accuracy and therefore are worthy of testing or additional measurements. The multivariate analysis is intended to establish the value of possible tests, such as a test to relate the coefficient of sliding friction between rail cars and the terrain. The committee is aware that the accura- cy of the multivariate regression likely will be limited by the lack of complete data on the broad varia- bility of derailments and that the merged database might lack detail on several of the key variables of interest. However, this procedure provides an objective approach in helping to ensure that DOT’s plan is focused on whether the most important assumptions are justified. Modeling and Simulation Approach Validation 3.1. Finding The ability of the modeling approach used for the HHFT final rule to simulate specific derailment events given a set of parameters was not adequately validated. Instead of examining train behavior dur- ing specific events, the validation process primarily compared simulated results for the number of cars derailed and the number of punctures against a large data scatter from historical derailment events. 3.1. Recommendation DOT should validate its modeling approach by comparing the model’s prediction of the outcome of an individual derailment event with the actual outcome of that one event. The type of braking system in- cluded in a modeled scenario should be the same as that involved in the actual train derailment with which the model prediction is to be compared. Such a comparison should be conducted for a reasona- bly large number of real-world derailments of diverse types. Adequate physics-based metrics should be the basis of comparisons between the event and the model predictions. The acquired merged database (see Recommendation 2.2) should be used for the validation tests. The physics-based metrics used (for example, the number of tank cars derailed or the number of tank cars punctured) will depend on the accident data available from the merged database. The use of scatter plots would not be appropriate for the recommended validation. Relevant examples of successful vali- dation efforts are those conducted on computer simulations used for the prediction of crash deformation of motor vehicles.59 59Ray, M. H., M. Mongiardini, C. A. Plaxico, and M. Anghileri. NCHRP Web-Only Document 179: Procedures for Verification and Validation of Computer Simulations Used for Roadside Safety Applications. Transportation Research Board of the National Academies, Washington, D.C., 2010. http://www.trb.org/main/blurbs/166054.aspx. Marzougui, D., C.-D. Kan, and K. S. Opiela. Comparison of Crash Test and Simulation Results for Impact of Sil- verado Pickup into New Jersey Barrier Under Manual for Assessing Safety Hardware. Transportation Research Rec- ord: Journal of the Transportation Research Board, No. 2309, 2012, pp. 114–126. http://trrjournalonline.trb.org/ doi/abs/10.3141/2309-12.

27 Appendix A Biographical Information: Committee on the Review of Department of Transportation Testing of Electronically Controlled Pneumatic Brakes Louis J. Lanzerotti, Chair, is Distinguished Research Professor in the Department of Physics at the New Jersey Institute of Technology. He is a retired Distinguished Member of Technical Staff of Bell Laboratories Lucent Technologies, where his responsibilities included supervision of laboratories and research and devel- opment. His principal research interests include space plasmas, geophysics, and engineering problems related to the impacts of atmospheric and space processes and the space environment on space and terrestrial tech- nologies. He has served as chair of a number of National Research Council boards and committees, including the Space Studies Board, the Committee on Electric Vehicle Controls and Unintended Acceleration, the Committee on the Assessment of Options for Extending the Life of the Hubble Space Telescope, and the Committee on Safety and Security of Commercial Spent Nuclear Fuel Storage. Dr. Lanzerotti also served as a member of the Academies’ Report Review Committee. He has been principal investigator (PI) on National Aeronautics and Space Administration (NASA) and commercial space satellite missions; he was PI for in- struments on the currently flying NASA dual spacecraft Van Allen Probes mission. Dr. Lanzerotti was elect- ed to the National Academy of Engineering in 1988. He received a PhD in physics from Harvard University. Mehdi Ahmadian is Dan Pletta Professor of Mechanical Engineering at Virginia Polytechnic Institute and State University (Virginia Tech), where he is also Director of the Center for Vehicle Systems and Safety (CVeSS) and the Railway Technologies Laboratory (RTL). He is the founding director of CVeSS, RTL, the Virginia Institute for Performance Engineering and Research, and the Advanced Vehicle Dynamics Labora- tory. Before joining Virginia Tech, Dr. Ahmadian worked as Lead Design Engineer at General Electric Transportation Systems and in various engineering and management positions at Lord Corporation. He has authored more than 130 archival journal publications and more than 250 conference publications and has given a number of keynote lectures. He has edited six technical proceedings and has made more than 300 technical presentations on topics related to advanced technologies for ground vehicles. He holds 10 U.S. and international patents and has edited four technical volumes. He serves as Editor for the International Journal of Vehicle System Dynamics, Editor-in-Chief for the Journal of Vibration and Control, and Editor-in-Chief of the journal Shock and Vibration. He has served as Editor-in-Chief for Advances in Automobile Engineer- ing (2010–2014) and as Associate Editor for the American Society of Mechanical Engineers Journal of Vi- bration and Acoustics (1989–1996), the American Institute of Aeronautics and Astronautics Journal (2000– 2008), and Shock and Vibration (2003–2011). Dr. Ahmadian is a Fellow of American Society of Mechanical Engineers, a Fellow of the Society of Automotive Engineers (SAE International), and an Associate Fellow of the American Institute for Aeronautics and Astronautics. His most recent professional awards include the 2008 SAE International Forest R. McFarland Award; the 2014 SAE International L. Ray Buckendale Award with a plenary lecture on Integrating Electromechanical Systems in Commercial Vehicles for Improved Han- dling, Stability, and Comfort; and the 2015 SAE Lloyd L. Withrow Distinguished Speaker Award. Dr. Ah- madian is an active member of SAE International. His activities include Member, Executive Nominating Committee; Member-at-Large, SAE Membership Board; Executive Council Member, SAE International Commercial Vehicle Engineering Congress (COMVEC); Activity Chair, COMVEC 2015; Activity Chair, COMVEC 2010; and past Chair of the SAE Chassis and Suspension Committee. He received a PhD in me- chanical engineering from the State University of New York at Buffalo.

28 Marshall Beck, of M. Beck Consulting, LLC, is the former Senior Vice President of Marketing and Sales at New York Air Brake Company (retired). During his 25-year career at New York Air Brake, he provided leadership on corporate vision and product innovation efforts, including the design and launch of the loco- motive computer control brake (CCB II), EP-60, and LEADER train-handling technologies. Mr. Beck holds a number of related patents. Before he joined New York Air Brake, he was the Director of Marketing at Bombardier, where he was responsible for passenger rail in both commuter and intercity markets. He has a master’s degree and a bachelor’s degree in mechanical engineering from the Royal Military College of Can- ada. Philip J. Daum is a Principal at Engineering Systems, Inc. He specializes in mechanical engineering, re- search, development, and experimental testing. He has conducted complex, multidisciplinary research and failure and accident investigations pertaining to freight and transit railroads, cargo and portable tanks, trans- portation equipment systems and components, and hazardous materials. Mr. Daum has investigated reliabil- ity, durability, crashworthiness, security performance, and regulatory compliance. His industrial experience over 35 years includes design of railroad rolling stock, trucks and bogies, brakes, draft systems, valves, pres- sure relief devices, safety equipment, and operating components and systems. He is also experienced in tech- nology evaluations, intellectual property analysis, equipment qualification, health monitoring, and mainte- nance. Mr. Daum received a BS in mechanical engineering from the University of Illinois at Urbana– Champaign and holds professional engineering licenses in Illinois and California. Jenny L. Ferren is the manager of the Structural Dynamics and Product Assurance Section at Southwest Research Institute (SwRI). She leads a team of engineers and technicians engaged in applied research, devel- opment, and independent testing services for the aerospace, automotive, nuclear, roadside safety, telecom- munications, marine, and oil and gas industries and for the U.S. military. She has more than 20 years of ex- perience in experimental structural dynamics, physical environmental testing, failure mode analysis, data acquisition, and analysis, as well as mechanical design and product development, for which she holds two patents. Her project work at SwRI has primarily included engineering stress screening and qualification test- ing to assess functionality, structural integrity, and compatibility of equipment subjected to extreme envi- ronmental conditions. She oversees reliability evaluations and model validation testing to support structural and dynamic analysis of large-scale systems. Ms. Ferren is responsible for business development efforts, in- cluding financial growth and promotional activities for the section, and serves as the system administrator responsible for maintaining the Structural Engineering Department’s ISO 9001, ISO 17025, and NQA-1 quality programs. She received a BS in mechanical engineering from Texas A&M University and an MBA from Texas A&M University–San Antonio. Roger L. McCarthy, of McCarthy Engineering, is a private engineering consultant and a director of Shui on Land, Ltd., which is involved in large-scale urban redevelopment in the People’s Republic of China. Dr. McCarthy has substantial experience in the analysis of failures of an engineering or scientific nature. He has investigated the grounding of the Exxon Valdez, the explosion and loss of the Piper Alpha oil platform in the North Sea, the fire and explosion on the semisubmersible Glomar Arctic II, and the rudder failure on the very large crude carrier Amoco Cadiz. Previously, Dr. McCarthy was chairman emeritus of Exponent, Inc., and chairman of Exponent Science and Technology Consulting Company, Ltd. (Hangzhou, China). In 1992, he was appointed by the first President Bush to the President’s Commission on the National Medal of Science. Dr. McCarthy served as a member of the National Academy of Engineering and National Research Council Committee on the Analysis of Causes of the Deepwater Horizon Explosion, Fire, and Oil Spill to Identify Measures to Prevent Similar Accidents in the Future and of the Committee on Options for Implementing the Requirement of Best Available and Safest Technologies for Offshore Oil and Gas Operations. He served as a member of the Committee on the Evaluation of the Federal Railroad Administration Research and Develop- ment Program. Dr. McCarthy received a PhD in mechanical engineering from Massachusetts Institute of Technology (MIT). He was elected to the National Academy of Engineering in 2004.

29 Raúl Radovitzky is Associate Director, Institute for Soldier Nanotechnologies, and a Professor of Aero- nautics and Astronautics, MIT. He joined MIT’s Department of Aeronautics and Astronautics in 2001 as the Charles Stark Draper Assistant Professor. His research interests are in the development of advanced concepts and material systems for blast, ballistic, and impact protection. To this end, his research group develops theo- retical and computational descriptions of the physical event and its effects on structures and humans, includ- ing advanced computational methods and algorithms for large-scale simulation. The resulting models help to improve understanding of the various physical components of the problem and thus to design protective sys- tems. Dr. Radovitzky’s educational interests include computational mechanics, continuum mechanics, aero- space structures, mechanics of materials, numerical methods, and high-performance computing. He is a member of the American Institute of Aeronautics and Astronautics, the International Association of Compu- tational Mechanics, the American Academy of Mechanics, the Materials Research Society, the U.S. Associa- tion of Computational Mechanics, and the American Society of Mechanical Engineers. He has served on the following National Academies committees: the Committee on Review of Test Protocols Used by the DoD to Test Combat Helmets and the Committee on Opportunities in Protection Materials Science and Technology for Future Army Applications. Dr. Radovitzky received a PhD in aeronautical engineering from the Califor- nia Institute of Technology. Patrick J. Student is former Director of Hazardous Material and Hazardous Materials Management, Union Pacific Railroad Company (retired). He has more than 41 years of experience in hazardous materials trans- portation by rail. He was responsible for interpreting hazardous materials regulations and railroad operating rules and developing systems for compliance with them. He was also a consultant on tank car issues and properties of materials being transported. Mr. Student has served as chairman and member of the AAR Haz- ardous Materials Committee and the AAR Tank Car Committee and has been chairman or a member of many of their task forces. He was a member of the Next Generation Rail Tank Car Project and is a member of the Advanced Tank Car Collaborative Research Project. He is a developer of structured electronic data interchange (EDI) for hazardous materials transportation and has served as chairman of the AAR EDI Haz- ardous Materials Technical Advisory Group. Mr. Student holds a BS in chemistry from the University of Missouri at Rolla. Gerhard A. Thelen is the former Vice President of Operations Planning and Support for Norfolk Southern Corporation, where he was responsible for operations planning, policies, budgets, research and tests, and quality management functions from 2006 until he retired in 2013. Mr. Thelen joined Norfolk Southern in 1977 and served in quality, engineering, research, and mechanical positions—including Assistant Vice Pres- ident of Research and Tests—before being named Vice President, Mechanical, in 2004. Norfolk Southern began running unit trains with electronically controlled pneumatic brakes on certain lines in late 2007. He served as a member of the National Academies Committee to Review USDOT Five-Year RD&T Strategic Plan and the Committee for Review of the Federal Railroad Administration Research, Development, and Demonstration Programs. Mr. Thelen has a master’s degree in industrial engineering from Pennsylvania State University. Elton Toma is a senior engineer at the Automotive and Surface Transportation portfolio at the National Re- search Council Canada. In that position he has worked on research projects related to the testing and analysis of all modes of surface transportation, including freight and passenger rail vehicles, heavy-haul highway trucks, and off-road military vehicles. Most recently, he has worked closely with Transport Canada on guide- lines for the marshaling of trains to reduce in-train forces, on the analysis of emergency stopping distances of freight trains, and on route risk analysis. Previously, he was a vehicle dynamics analyst at the Transportation Technology Center, Inc. (TTCI), in Pueblo, Colorado. At TTCI he was involved in the on-track testing and computer modeling of freight and passenger rail vehicles. He has a PhD in mechanical engineering from Queen’s University. His PhD research focused on the analysis of train derailments. His work produced a computer model of a train derailment, which was the only model available for derailment simulation at the time.

30 Appendix B Statement of Task In the first phase of this project, an ad hoc committee will review a test and analysis plan prepared by DOT and comment in a letter report on whether the proposed tests will provide objective, accurate, and reli- able results to test the assumptions that DOT has identified in its comparison of the emergency braking per- formance of railroad tank car electronically controlled pneumatic (ECP) brakes to conventional brakes or braking systems such as distributed power and two-way end-of-train devices. The committee will provide a written explanation detailing the need for any additional or alternative testing. The key question is whether ECP brakes would reduce the incidence and severity of spills of crude oil or ethanol from derailments com- pared with the alternative braking systems examined. In the second phase of this project, the committee will review the conduct of DOT’s tests, reports of test results, and, based on DOT’s test results and analysis, provide its findings and conclusions addressing the performance of ECP brakes relative to other braking technologies or systems tested by DOT. The com- mittee’s reviews and letter reports will be limited to these tasks; the committee will not make recommenda- tions about which braking systems should be required of railroads in revenue service.

31 Appendix C Information-Gathering Activities of the Committee In the course of preparing this report, the committee met four times. At one of those meetings, oral presentations were made by the following individuals in public session at the invitation of the committee: Kevin Kesler, FRA, DOT; Joe Brosseau, Transportation Technology Center, Inc.; N. Scott Murray, Exxon- Mobil; Mike Parisian, NYAB; and John Halowell, Wabtec. The committee reviewed various written materi- als from NTSB, AAR, and others. In addition, the committee reviewed test rack data it had requested from NYAB and Wabtec. Interested members of the public at large were given an opportunity to speak to the committee. In accordance with institutional procedures, the committee visited the FRA test facility in Pueblo, Colo- rado, on September 28, 2016, to observe a tank car test that was part of implementing the test plan (see Ap- pendix E). FRA indicated that it began to implement the test plan because of the short amount of time avail- able for completing the steps needed to inform the Secretary of Transportation’s decision that was called for by Congress in the FAST Act. On October 25, 2016, the committee chair sent a letter to Secretary Foxx expressing concern that con- ducting additional tests without consideration of the committee’s advice might result in the inefficient use of limited resources. In that letter, the committee recommended that DOT suspend its testing activities until the committee’s report had been issued.1 1The letter is available at http://www.trb.org/PolicyStudies/ReviewofECPBrakes.aspx.

32 Appendix D Acknowledgment of Reviewers This report has been reviewed in draft form by persons chosen for their diverse perspectives and tech- nical expertise in accordance with procedures approved by the NASEM Report Review Committee. The pur- poses of the independent review are to provide candid and critical comments that will assist the institution in making its published report as sound as possible and to ensure that the report meets institutional standards of objectivity, evidence, and responsiveness to the study charge. The review comments and draft manuscript remain confidential to protect the integrity of the deliberative process. The committee thanks the following individuals for their review of this report: Christopher Barkan, University of Illinois at Urbana–Champaign; John Halowell, Wabtec Corporation; Steve Kirkpatrick, Applied Research Associates, Inc.; Xiang Liu, Rut- gers, The State University of New Jersey; Bryan McLaughlin, New York Air Brake; Robert McMeeking (NAE), University of California, Santa Barbara; Larry Milhon, BNSF Railway Company (retired); Ed Pritchard, Paladin Consulting Group, Inc. (retired); Kevin J. Renze, consultant; Walter Rosenberger, Norfolk Southern Corporation; John M. Samuels, Jr. (NAE), Revenue Variable Engineering, LLC; Roger R. Schmidt (NAE), International Business Machines Corporation; and Terry Tse, Federal Railroad Administration (re- tired). Although the reviewers listed above provided many constructive comments and suggestions, they were not asked to endorse the conclusions or recommendations, nor did they see the final draft of the report before its release. The review of the report was overseen by the review coordinator, Susan Hanson (NAS), Clark University (emerita), and the review monitor, Robert Sproull (NAE), University of Massachusetts at Am- herst. Appointed by NASEM, they were responsible for making certain that an independent examination of the report was carried out in accordance with institutional procedures and that all review comments were carefully considered. Responsibility for the final content of the report rests entirely with the committee and the institution.