David Jeong

This report describes research conducted to examine the application of sandwich structure technology to provide protection against the threat of an indenter striking the side or shell of a tank car in the event of an accident. This research was conducted in two phases over a 3-year period. Testing and analysis of flat, welded steel sandwich panels was conducted in the initial phase of the research. Based on the observations and results from that initial phase, a curved, welded steel sandwich panel was designed and built to protect the side or shell of a decommissioned liquid chlorine tank car during a full-scale impact test. Although the protective panel experienced severe damage, the commodity-carrying tank experienced only permanent deformation and did not puncture.

This paper describes analyses of a railroad tank car impacted at its side by a ram car with a rigid punch. This generalized collision, referred to as a shell impact, is examined using nonlinear (i.e., elastic-plastic) finite element analysis (FEA) and three-dimensional (3-D) collision dynamics modeling. Moreover, the analysis results are compared to full-scale test data to validate the models. Commercial software packages are used to carry out the nonlinear FEA (ABAQUS and LS-DYNA) and the 3-D collision dynamics analysis (ADAMS). Model results from the two finite element codes are compared to verify the analysis methodology. Results from static, nonlinear FEA are compared to closed-form solutions based on rigid-plastic collapse for additional verification of the analysis. Results from dynamic, nonlinear FEA are compared to data obtained from full-scale tests to validate the analysis. The collision dynamics model is calibrated using test data. While the nonlinear FEA requires high co...

The most common rail defect encountered in continuously welded rail is known as the detail fracture. The U.S. Department of Transportation, Federal Railroad Administration has sponsored and managed research over the past several decades to understand the structural integrity of rail in general, and the fatigue crack growth behavior of detail fractures in particular. Control of rail integrity and defect growth is conducted via periodic rail tests (i.e. inspections) to ensure that rail defects do not become large enough to cause rail failure. Moreover, federal regulations have been codified to establish a maximum interval between rail inspections based on the results of government-sponsored research. Over the past several decades, however, rail manufacturing has evolved and improved, particularly the head-hardening process to improve wear resistance. Propagation life of railroad rail was examined in previous research using fatigue crack growth data for non-head-hardened rail. Recently Thornton-Tomasetti conducted research, sponsored by FRA, to examine the fatigue crack growth behavior of modern rail steels (i.e. railroad rails with head-hardening). The initial results of the more recent research effort were reported in the 2019 Joint Rail Conference. In this paper, fatigue crack growth rate data generated for head-hardened rail are used to examine the fatigue crack growth life of detail fractures under nominal revenue service conditions. Moreover, this paper applies a probabilistic approach to estimate rail life to account for the inherent variability or scatter typically observed in fatigue crack growth rate data. Regression methods are employed to derive the parameters for the Walker crack growth rate equation, which are subsequently treated as correlated, multivariate, and normally distributed random variables. Data from four different rail steels are used in the regression analyses, which are referred to as: Advanced Head Hardened (AHH), Head Hardened (HH), Standard Strength (SS), and Colorado Fuel and Iron (CF&I). Monte Carlo simulations of fatigue growth of detail fractures are carried out to estimate fatigue life distributions for each of the different rails. The results from these four rail steels are compared to those based on the previous research for non-head-hardened rails. Implications of these comparisons on determining rail testing intervals are discussed.

The Federal Railroad Administration (FRA) has been sponsoring research on rail integrity for several decades. This research has been chiefly managed and conducted by the Volpe National Transportation Systems Center (Volpe). Particular focus has been given in this research to rail head defects, known as detail fractures, since they are the most commonly encountered defect in continuous welded rail track [1]. Testing and analyses have been performed on railroad rails manufactured without head hardening. Modern rail, however, are now heat treated during the manufacturing process to harden the rail surface to increase its resistance to wear. As such, the heat treatment and nonuniform cooling induce complex residual stress patterns in the rail that can affect microstructure and fatigue crack growth rate behavior. This paper will describe research to examine defect growth behavior of modern rail steels. This research is a collaboration among several organizations: Thornton-Tomasetti, Arcelor-Mittal, Lehigh University, Harvard University, National Institute of Standards and Technology (NIST), Fraunhofer Institute, and Volpe. Arcelor Mittal donated rails with different grades of steel: advanced head hardened, head hardened, and standard strength (i.e. non-head-hardened). Lehigh conducted laboratory tests on specimens cut from these rails to perform various tests, which include: hardness measurements, mechanical testing to measure tensile properties, fracture toughness measurements, and fatigue crack growth rate tests. All of these tests were performed in accordance with applicable ASTM International standards. NIST and Fraunhofer performed preliminary neutron diffraction measurements of residual stresses on the different rails. Moreover, this paper will present results from the laboratory testing program. Implications of these results on detail fracture growth behavior will also be discussed.

A series of tests, aimed at assessing the structural integrity of joint bars under differing service conditions, were conducted to address concerns regarding joint bar failures in the revenue service environment. Data collected through the course of this study revealed that bending stress invoked by normal track surfacing operations is not a likely cause for cracks that initiate at the top center of joint bars. Instead, cracking at this location is probably the result of fatigue at the top center of the joint bar due to rail-joint contact. Surface hardening at the area of rail-joint contact was largely ineffective, resulting in metal flow developing adjacent to the easement at the top of the joint bar. Additional data gathered in this study suggests little correlation between the surface hardness of the joint bar and the depth of metal flow. Bending stresses and wheel/rail forces were also measured on joint bars used in rail end gaps and rail height mismatches, which revealed minima...

Indented wires have been increasingly employed by concrete crosstie manufacturers to improve the bond between prestressing steel reinforcements and concrete, as bond can affect several critical performance measures, including transfer length, splitting propensity and flexural moment capacity of concrete ties. While extensive experimental testing has been conducted at Kansas State University (KSU) to obtain bond characteristics of about a dozen commonly used prestressing wires, this paper develops macro-scale or phenomenological finite element bond models for three typical wires with spiral or chevron indent patterns. The steel wire-concrete interface is homogenized and represented with a thin layer of cohesive elements sandwiched between steel and concrete elements. The cohesive elements are assigned traction-displacement constitutive or bond relations that are defined in terms of normal and shear stresses versus interfacial dilatation and slip within the elasto-plastic framework. A...

... Hailing Yu1, David Jeong, John Choros and Ted Sussmann Structures and Dynamics Division John A. Volpe National Transportation Systems ... adjacent ties, dynamic forces and rail pad attenuation [4]. Dynamic forces may result from irregularities on wheel and rail surfaces ...

This paper describes nonlinear finite element analysis (FEA) to examine the energy to fracture unnotched Charpy specimens under pendulum impact loading. An oversized, nonstandard pendulum impactor, called the Bulk Fracture Charpy Machine (BFCM), was constructed to conduct studies to assess the fracture behavior of various tank car steels. Comparisons between measured and calculated impact energies are presented. The effect of various factors on impact energy is demonstrated through the test data and FEA results. These factors include striker size and shape, specimen thickness, and specimen material. Moreover, calculations of fracture energy are carried out using FEA in conjunction with different material failure criteria. The FEA results using a failure initiation criterion based on the general state of stress in terms of stress triaxiality and a failure progression model based on linear strain softening are shown to provide excellent agreement with the experimental data.

ABSTRACT When a polycrystal is stressed or strained at fifty percent of the corresponding yield value, damage will be inflicted non-homogeneously in the material due to the fact that the stress and/or strain distribution is non-uniform even if isotropy and homogeneity are assumed for the initial microstructure. This effect will be cumulated for each cycle of the load if the applied stress or strain is repeated continuously. Nucleation of microcracks can eventually lead to the propagation of a macrocrack.The process of damage accumulation in fatigue is defined to be sufficiently slow such that inhomogeneity of material behavior created by loading is a significant factor that can not be arbitrarily dismissed without a good reason. What this means specifically is that the difference of the stress and strain behavior at each point in a fatigue specimen must be accounted for in the analytical model in order to predict the correct cumulative effect. Such a requirement translates into a non-equilibrium formulation where the constitutive relations for each point and loading cycle must be determined separately. In this sense, the true problem of fatigue cannot be completely treated by the classical continuum mechanics approach that is limited to equilibrium mechanics for a closed system. Having said this, the isoenergy density theory will be applied to estimate the hysteresis loops of a hour-glass profile cylindrical bar specimen as recommended by the American Society for Testing and Materials (ASTM) for low-cycle fatigue.The work will be divided into two parts. Part I will cover the fundamentals of a non-equilibrium theory where the continuum elements are finite in size; they do not vanish in the limit. Therefore, size effects are immediately encountered as a function of time. General expressions for the rate change of volume of these elements with surface area are derived such that they can be computed from the nine displacement gradients. These elements can differ in size and must fit together without discontinuities or gaps to form the continuum. The condition of isoenergy energy density is invoked such that the size of these individual elements under large and finite deformation and rotation can be determined without loss in generality. The existence of such a space having the property of the same isoenergy density in all directions is thus proved. This enables the establishment of the one dimensional energy state with that in three dimensions without restriction, the absence of which has prevented the development of a complete non-linear theory of mechanics that can be solved in a direct fashion in contrast to the inverse method of assuming the displacement field. Illustration is provided for deriving the constitutive relation incrementally for a given location for the hour-glass specimen made of 6061-T6 aluminum. Once the specimen is loaded, each material point will follow a different stress and strain curve according to the local displacement rate. Hence, the method applies to material with non-homogeneous microstructure if their individual expressions can be assessed and fed into the computer.Part II computes for the non-equilibrium temperature and an entropy-like quantity that can be positive and negative. This implies that the system can absorb or dissipate energy with reference to the surrounding. Additional data for hysteresis loops are given for 6061-T6 aluminum, SAE 4340 steel and Ti–8Al–1Mo–1V titanium. Accumulation of the local hysteresis energy per cycle is found to be the highest near the surface of the uniaxial specimen where load symmetry prevails. This is a consequence of the difference in accumulation of the energy density due to distortion in contrast to dilatation at the specimen center. This is why fatigue cracks tend to nucleate near the specimen surface, at a small distance towards the interior. Another distinct feature of fatigue is that the non-equilibrium temperature is found to oscillate about the ambient temperature while the local stress states fluctuate between tension and compression. This temperature reversal behavior is typical of non-equilibrium behavior and also occurs under monotonic loading. The space and time variations of the dissipated energy density for different materials are found to be related to the initial monotonic energy density or area under the true stress and true strain curve.What will be demonstrated is that no special consideration need to be made when applying the isoenergy density theory for analyzing the nucleation of micro and macrocracks in addition to failure of the specimen. Crack nucleation under fatigue is assumed to occur when the total hysteresis energy reaches a critical value. It is possible to establish a relation between the average hysteresis energy per cycle and the number of cycles to failure. The proposed method requires only a knowledge of the initial monotonic energy density curve for a given material. Predicted results for the fatigue…

Several industries now use risk analysis to develop inspection programs to ensure acceptable mechanical integrity and reliability. These industries include nuclear and electric power generation, oil refining, gas processing, onshore and offshore exploration and production, chemical processing, and pipelines. Risk analysis may also be used as a decision-making tool in the railroad industry to develop systematic improvements in track maintenance and inspection strategies. In the course of conducting research in support of the Federal Railroad Administration, a Monte Carlo risk assessment model has been developed to simulate certain aspects of rail inspection (also referred to as rail testing) to find and remove defects that may grow to sufficient size to cause rail failures. In this paper, the model is used to examine the relationship between the occurrence of rail failures and various operational factors. These operational factors include rail size, average axle loading, and inspecti...

Abstract: In this article the authors discuss research into the stability of bolted joint bars as necessitated by the disastrous consequences of joint bar failure such as that of the Minot, North Dakota failure in 2002. Using a three-dimensional finite element analysis, bending ...

This paper describes engineering analyses of a railroad tank car impacted at its head by a rigid punch. This type of collision, referred to as a head impact, is examined using dynamic, nonlinear finite element analysis (FEA). Commercial software packages ABAQUS and LS-DYNA are used to carry out the nonlinear FEA. The sloshing response of fluid and coupled dynamic behavior between the fluid inside the tank car and the tank structure are characterized in the model using both Lagrangian and Eulerian mesh formulations. The analyses are applied to examine the structural behavior of railroad tank cars under a generalized head impact scenario. Structural behavior is calculated in terms of forces, deformations, and puncture resistance. Results from the two finite element codes are compared to verify this methodology for head impacts. In addition, FEA results are compared to those from a semi-empirical method.

ABSTRACT This paper describes analyses to examine the lateral deflection of railroad track subjected to quasi-static loading. Rails are assumed to behave as beams in bending. Movement of the track in the lateral plane is constrained by idealized resistance characteristics, while movement in the vertical plane is resisted by a continuous, linear and elastic foundation. These analyses are based on solving the ordinary differential equations for beam deflections. In certain cases, convenient mathematical expressions may be used to represent idealized lateral resistance characteristics and derive closed-form equations to relate lateral force as a function of track lateral deflection. However, in general, the idealized lateral resistance characteristic may be nonlinear, in which case numerical methods are required to examine the lateral load versus track lateral deflection behavior. In these general cases, a Fourier series technique is used to solve the governing equations numerically.The analysis of track lateral deflection subjected to quasi-static loads may be applied to examine track shift. For example, lateral resistance of track may be measured using Track Lateral Pull Tests (TLPT). The Fourier method is also used to examine the relationship between lateral and vertical wheel loads and track lateral shift.

In this paper, a nonlinear bulging factor is derived using a strain energy approach combined with dimensional analysis. The functional form of the bulging factor contains an empirical constant that is determined using R-curve data from unstiffened flat and curved panel tests. The determination of this empirical constant is based on the assumption that the R-curve is the same for both flat and curved panels.

A ballasted railroad track consists of rails, fasteners, ties, ballast and the underlying subgrade. Realistic simulations of the track response to interactive vehicle-rail loads will require detailed models of each component. In this paper, finite element (FE) models are developed for some of the ballasted track components including wood and concrete crossties, ballast and subgrade. A user material subroutine is employed to predict the failure of wood ties based on an orthotropic stress criterion. The concrete tie is modeled as a heterogeneous medium with prestressing tendons embedded in a concrete matrix. The concrete material observes elasticity followed by damaged plasticity. The interfaces between the tendons and concrete are simulated with cohesive elements, and there can be initiation and evolution of damage to the bond between the steel tendons and concrete. Further, the granular and frictional ballast material is modeled with extended Drucker-Prager plasticity, and the subgr...

In support of the Federal Aviation Administration Technical Center's (FMTC) National Aging Aircraft Research Program (NAARP), Sandia National Laboratories and the John A. Volpe National Transportation Systems Center (Volpe Center) are conducting research to determine if current rules for design, inspection, and maintenance are sufficient to ensure the safe operation of the aging fleet. Particular emphasis has been given to a phenomenon of multiple cracking that appears to be an attribute of airplanes that have been in service for some time. This phenomenon is commonly referred to as Widespread Fatigue Damage (WFD). Several experimental and analytical studies have been initiated by FAATC to understand the phenomenon of WFD. Some of these research activities include: collection of strain gage data from a Boeing 737 airplane conducted by the Aging Aircraft Nondestructive Inspection Validation Center (AANC); laboratory testing of full-scale curved panels conducted by Foster-Miller, ...

This paper presents a computational framework that employs elastic-plastic-failure finite element analysis (FEA) tools to predict dynamic forces, deformations and puncture resistance of railroad tank car heads in impact events. First, computational sensitivity to analysis code, analysis type, element type, element size, through-the-thickness characteristics and element integration scheme is studied for elastic-plastic analyses. The fully integrated shell element formulation in ABAQUS/Explicit yields acceptable global impact force-indentation responses in dynamic analyses, and the results converge with 3-4 through-the-thickness Gaussian integration points and a characteristic element size 1-2 times the tank thickness. Second, the progressive damage and failure modeling for ductile metals is employed to estimate the puncture resistance of a tank car head. The Bao-Wierzbicki fracture initiation criterion expressed in the stress triaxiality-equivalent plastic strain plane is employed to...

ABSTRACT In May 2011, a derailment of a passenger train occurred in a tunnel in the northeast region of the United States. Fortunately, no serious injuries or fatalities resulted from this derailment. The probable cause of the derailment was determined to be a broken rail from a defect originating in the base of the rail. This internal rail base defect is characterized as having a crescent, thumbnail, or semi-elliptical shape. In addition, the formation and growth of this defect may have been exacerbated by corrosion.This paper describes engineering calculations to estimate the growth rate of this type of rail base defect. These engineering calculations are based on applying the principles of fracture mechanics and beam theory. Fracture mechanics principles are applied to determine stress intensity factors for the semi-elliptical shaped defect with different aspect ratios. Stress intensity factors are then used to estimate the growth of the defect under the accumulation of tonnage from repeated wheel passages. For this purpose, the rail is assumed to behave as a beam in bending.

ABSTRACT Railroad tank cars are exposed to a high degree of variable amplitude loading that is comprised of both tensile compressive cycles. In order to consistently and accurately predict this type of loading, material properties, usage, and load interaction models must be used in concert. The objective of this study is to present the methodology necessary to generate tank car specific spectrum crack growth life prediction models. Experimental fatigue crack growth data from this project and previous efforts provided the baseline data necessary to develop NASGRO® models and predict simplified over-load and under-load experiments. The methods were then advanced to predict spectrum crack growth experimental data representative of tank car usage. The results show that significant amounts of compressive loads in the tank car spectra essentially negate the need for a tensile stress based load interaction model. In order to provide consistently conservative prediction, the most basic model must be implemented instead of the more elaborate retardation models.