This paper provides an introduction to a Rail Break Detection and Train Location Detection using fiber optic technology called Smart Rail. The technology concept detailed in this paper not only detects the location of a broken rail, it also provides advanced warning of rail damage that may evolve into a dangerous condition. The same technology allows for the detection of trains and vehicles in the detection zone and provides location, speed and direction information. The Smart Rail continuously monitors rails for changes in their ability support the train in a safe manner. Smart Rail technology is based on direct observations of the structural quality of the rail by means of rail strain detection. Smart Rail technology detects the location, speed and direction of vehicles on the rails without the need for onboard equipment. It reports the location of vehicles whether moving or not. Smart Rail technology is also capable of detecting and providing a warning of: rail gauge separation, lose or missing track hardware, missing tie plates and flat wheels. Taken together this technology can assist in the prevention of derailments. The technology can be applied to existing rails with no need to modify any existing equipment or track circuits. The train detection range can be up to 50 miles per direction from one wayside location.

Engineered polymer composite (EPC) ties offer a potential alternative to solid sawn timber ties. These materials are especially attractive for use in regions where wood is susceptible to degradation by moisture and decay organisms. However, recent research at the Transportation Technology Center’s (TTC) Facility for Accelerated Service Testing (FAST) in Pueblo, CO, found that track supported by EPC ties experienced more gage widening variation due to temperature changes than track supported by wood ties. Specifically, the track gage was about 0.2-in. wider in the afternoon than that in the morning on the EPC tie tracks. It is believed that the direct sunlight in the afternoon makes the top surface of the tie expand more than the other parts of the tie, thereby causing the EPC ties to bend and widen track gage. Another observation related to the EPC thermal bending effect is changes to the ballast support condition. When temperatures are cooler, EPC ties tend to experience a center-bound ballast support condition, therefore generating more bending stress on the ties. This paper presents results from computer simulations of the thermal behavior of EPC ties. Future study will focus on field testing to further understand the thermal effects in support of recommendations on the use of EPC ties.

Cut spike fasteners, used with conventional AREMA rolled tie plates and solid sawn timber ties, are the most common tie and fastener system used on North American freight railroads. Cut spikes are also used to restrain tie plates that incorporate an elastic rail fastener — that is, an elastic clip that fastens the rail to the tie plate. Elastic fasteners have been shown to reduce gage widening and decrease the potential for rail roll compared to cut spike-only systems. For this reason, elastic fastener systems have been installed in high degree curves on many railroads. Recent observations on one Class I railroad have noted broken cut spikes when used with these types of tie plates in mountainous, high degree curve territory. Broken screw spikes and drive spikes on similar style plates have also been observed. In this paper, a simulation method that integrates a vehicle-track system dynamics model, NUCARS ® , with a finite element analysis model is used to investigate the root causes of the broken spikes. The NUCARS model consists of a detailed multibody train, wheel-rail contact parameters, and track model that can estimate the dynamic loading environment of the fastening system. For operating conditions in tangent and curve track, this loading environment is then replicated in a finite element model of the track structure — ties, tie plates, and cut spikes. The stress contours of the cut spikes generated in these simulations are compared to how cut spikes have failed in revenue service. The tuning and characterization of both the vehicle dynamics multibody model and the finite element models are presented. Additionally, the application of this approach to other types of fastening systems and spike types is discussed. Preliminary results have identified a mechanism involving the dynamic unloading of the tie plate-to-tie interface due to rail uplift ahead of the wheel and the resulting transfer of net longitudinal and lateral forces into the cut spikes. Continued analysis will attempt to confirm this mechanism and will focus on the severity of these stresses, the effect of increased grade, longitudinal train dynamics, braking forces, and curvature.

The Brazilian Rail Regulatory Framework was reviewed in 2011 by the National In Land Transport Agency (ANTT), with the edition of a package of Regulatory Resolutions (Resolution nº. 3,694 - Regulation of General Rules for Shippers and Rail Companies, Resolution nº. 3,695 - Regulation of Trackage and Haulage Rights, and Resolution nº. 3,696 - Regulation for Stretch Production Targets and Safety Targets Procedure Agreements) with the basic objectives of: i) create incentives for rail interoperability, ii) promote intra sector competition and, hence, enhance rail role in the national transport sector, and iii) establish new funding sources for the rail sector, in order to support investments needed for increasing rail system capacity. These, together with other structural measures already adopted by ANTT, such as the regulation of investments made by rail concessionaires and the revision of rail price cap rates, complemented the package of rail regulatory measures. For the effective implementation of the principles set out in the mentioned rail regulatory framework review, however, it was necessary to develop a sort of tools to allow regulatory control and reduce asymmetric information within the actors of this important industry. Among these tools, it is highlighted the Rail Network Statement, established by Resolution ANTT nº. 3,695/2011, basically composed of a compilation of the existing rail network main operational and technical data, such as rail gauges, load capacities, operational speeds, clearances, location of locomotives and wagons repair facilities as well as fuel stations, including an inventory of traffic capacity, with a survey of available capacity in all sections of the national rail network. This dataset was intended to reduce asymmetric information between rail industry players - Regulatory Authority, Rail Carriers and Shippers, and, so, enhance the negotiation processes for interoperability operations and investments agreements. Rail Network Statement must be filed annually by Concessionaires, and after a critical analysis and data review by the regulatory agency, it is released on its website for public access to support rail interoperability planning and negotiation process. This work is supposed to present an overview of Brazilian Rail Network Statement structure, with a Case Study Analysis of 2013 Edition, the first issued, with a strong focus on rail capacity analysis, an essential requirement for rail interoperability.

Accurate knowledge of transfer length has been shown to be crucial to the goal of maintaining continuous production quality in the modern manufacture of prestressed concrete railroad ties. Traditional manual laboratory methods, such as the conventional Whittemore method which requires the use of embedded reference points, are clearly not suitable for production operation or for use in reliable production quality-control. This paper presents the results of another advance in the development of automated transfer length measurement systems for practical in-plant operation. The new device offers a significant improvement over the previously successful automated Laser-Speckle Imaging (LSI) system developed by the authors. The earlier automated LSI strain measurement system has been modified to provide significantly improved optical resolution of longitudinal surface strain, with the ability to resolve longitudinal prestressed concrete crosstie surface strain without time-consuming special surface preparation. More importantly, the new system is also capable of making measurements of strain in a real-time “on-the-fly” manner over the entire distance range of interest on the tie associated with transfer length development. It features both a “jog” mode of operation, similar to its predecessor in which measurements of longitudinal surface strain are automatically captured in arbitrary spatial increments over the entire range of the computer-controlled traverse, and an “on-the-fly” mode in which measurements of longitudinal surface strain are captured without the need for stopping at each measurement location. This latter mode offers the potential of a much faster capture of the strain profile and should prove to be very beneficial for field testing and in-plant diagnostic applications. The performance of this new system is first demonstrated using a new calibrated step-wise uniform strain field setup which has been developed specifically for verification of this and other automated transfer length measurement systems. This verification system produces a calibrated step change in surface deflection, effectively subjecting the automated strain measurement system to an ideal step change in longitudinal strain for a given gauge length. In addition, the new automated system is demonstrated by conducting measurements of longitudinal surface strain on prestressed concrete crossties in a manufacturing plant. For this latter experimental in-plant testing, strain measurements using the new system are also compared directly with those from the recently introduced 6-camera transfer length measurement system, as well as with the traditional Whittemore gauge measurements. The agreement between these independent measurement systems is remarkable, and it is shown to even be possible to discern differences in strain profile and associated transfer length between adjacent crossties within a given casting bed. This new automated and high-resolution device should provide a very convenient and fast diagnostic tool for the manufacturer to quickly identify the need to modify production (e.g., concrete mix) if transfer length specifications fall out of desired range.

Railroad bridges experience dynamic wheel load augment from rolling stock that cross the bridges, due to nominal bridge and suspension dynamics, as well as anomalies such as wheel flats. The level of dynamic augment is particularly high for steam locomotives due to the hammer blow effect associated with the driven wheels. Tests conducted in the early-mid 20 th century had quantified some of these effects, and the resulting findings have been part of the impact formulae presented in the AREMA Railway Engineering Manual. However, there was concern that the impact associated with some of the non-cross counter-balanced, lighter, older locomotives, could be higher than specified by AREMA formulae. This paper describes the methodology and results from a series of tests that evaluated the levels of dynamic augment experienced by railroad track and an exemplar bridge under a set of narrow gauge steam locomotives, and compares the measurements to the design values specified in the AREMA Manual. Vertical and lateral loads on railroad track, and strain levels on multiple critical bridge members were measured under three different classes of light, narrow gauge steam locomotives, over a range of operating speeds and conditions. The tests were conducted on a 120 ft span, through truss bridge, and adjacent track on a tourist railroad. Dynamic augment values measured during the tests were generally lower than the values expected from AREMA formulae. Similarly, the peak lateral loads measured appear to be nominal and lower than the AREMA prescribed values. However, it should be kept in mind that these results are from tests conducted with three relatively light, narrow gauge locomotives, on specific bridge and track, whereas, the AREMA formulae are intended to cover a wider range of conditions. These tests tend to show that the legacy standards are conservative and are applicable to calculating regulatory required bridge loads where steam locomotives are concerned.

Transition zones in railway tracks are locations with considerable changes in the vertical stiffness of the rail support. Typically they are located near engineering structures, such as bridges, culverts and tunnels. In such locations, the vertical stiffness of the track support varies, resulting in amplification of the dynamic forces acting on the track, which ultimately leads to deterioration of the vertical track geometry. Also, differential settlement of the track sub-structure on both sides of the transition contributes to the deterioration of the vertical geometry. The deterioration process accelerates with increase of the operational velocities of the passing trains. Finally, all these result in increase of the maintenance efforts on correction of the track geometry in the transition zones. To analyze the dynamic behavior of transition zones due to the differential settlement, a finite element dynamic model (using explicit integration) of a track transition zone is developed. The model is also accounting for the effect of hanging sleepers. The model is verified against the measurement results performed using the Video Gauge System (VGS). With the developed model, the differential settlement of ballast and soil has been introduced and the effects of this settlement on performance of the track are analyzed. The resulting wheel forces and the dynamic responses of the track components are obtained and analyzed. Special attention has been paid on the stresses of ballast, which is one of the most vulnerable track components in the transition zones. Finally, conclusions on the effect of the various internal and external factors on the degradation process of the track in transition zones are drawn. With the developed model, the differential settlement of ballast and soil is induced and hence, the effects of the differential settlements can be analysed. The contact forces of wheel-rail interaction and the dynamic response of track components are obtained and analysed. Special attention has been paid on the stresses of ballast which are one of the most vulnerable components in the track transition zones. Finally, conclusions on the effect of the various internal and external factors on the degradation process of the track in transition zones are drawn.

This study is aimed to develop a real-time safety monitoring of kilometer-long joint rails using a distributed fiber optic sensor. The sensor measures the distribution of Brillouin frequency shift along its length with pulse pre-pump Brillouin optical time domain analysis (PPP-BOTDA). The measurement distance and spatial resolution can be up to 25 km and 2 cm, respectively. The fiber optic sensor was first characterized and calibrated for distributed strain and temperature measurement, and then instrumented on a small-scale joint rail-like specimen in laboratory. The specimen was loaded at room temperature, and its strain distribution along the sensor was measured using a Neubrescope with high accuracy and spatial resolution. Given a gage length, the joint open change was determined and visibly identified from the measured strain distribution. Finally, an implementation plan of distributed sensors on a railway is introduced, including sensor deployment, sensor repair when broken, and cost analysis. The gage length at a crack is an important parameter in sensor deployment and investigated using finite element analysis. The results indicate that the distributed sensor can be used successfully to monitor the strain and temperature distributions in joint rails.

The applicability of Doppler based, Light Detection and Ranging (LIDAR) sensors for track geometry monitoring is investigated in this article. The optimum configuration of the LIDAR lenses, suitable for effective lateral track geometry measurements, is presented. In the proposed embodiment, two low-elevation LIDAR optics are employed with their beams reflected off of the rail gage face on each side to capture speed signals from both rails individually. A processing technique is developed based on the frequency bandwidth dissimilarities between the vehicle speed and track geometry irregularity to separate the two from each other. LIDAR system is body-mounted to a geometry inspection car and tests are carried out on revenue service track. The results indicate that LIDAR optics can provide a reliable track monitoring instrument for use over substantial track mileage in inclement weather and harsh track conditions with minimal operator supervision.

As higher demands are placed on North American railroad infrastructure by heavy haul traffic, it is increasingly important to understand the factors affecting the magnitude and distribution of load imparted to concrete crosstie rail seats. The rail seat load distribution is critical to the analysis of failure mechanisms associated with rail seat deterioration (RSD), the degradation of the concrete surface at the crosstie rail seat. RSD can lead to wide gauge, cant deficiency, and an increased risk of rail rollover, and is therefore of primary concern to Class I Freight Railroads in North America. Researchers at the University of Illinois at Urbana-Champaign (UIUC) have successfully characterized the loading environment at the rail seat using matrix-based tactile surface sensors (MBTSS). Previous research has proven the feasibility of using MBTSS in both laboratory and field applications, and recent field experimentation has yielded several hypotheses concerning the effect of fastening system wear on the rail seat load distribution. This paper will focus on the analysis of data gathered from laboratory experimentation with MBTSS to evaluate these hypotheses, and will propose a metric for crosstie and fastening system design which considers the uniformity of the load distribution. The knowledge gained from this experimentation will be integrated with associated research conducted at UIUC to form the framework for a mechanistic design approach for concrete crossties and fastening systems.

Knowledge of transfer length is critical for maintaining continuous production quality in the modern manufacture of prestressed concrete railroad ties. Traditional manual laboratory methods for measuring transfer length are simply not suitable for production operation. They are much too time-consuming to implement, and typically require extensive surface preparation in order to obtain the required surface strain measurements needed for determining the transfer length. In addition, the traditional 95% Average Maximum Strain (95% AMS) method of assessing transfer length from the measured surface strain profile has been shown to possess bias. The accuracy of transfer length assessment using this method is generally influenced by individual operator judgment; therefore, making it unsuitable for use as a reliable production quality-control parameter. This paper presents recent in-plant testing of a newly developed prototype multi-camera non-contact transfer length measurement system, representing a major improvement over the traditional manual methods. The testing was conducted on concrete railroad ties at a manufacturing facility in North America. Concrete ties tested included those which were manufactured using indented prestressing wire as well as with 7-wire strand. The new device represents a next generation version of the previously successful Laser-Speckle Imaging (LSI) system developed by the authors. The multi-camera system is shown to provide nearly real-time surface strain profile measurements (subsequent to de-tensioning), with surface strain accuracy comparable to the mechanical Whittemore gage device, yet with little or no required surface preparation. Furthermore, with the previously demonstrated Zhao-Lee (ZL) transfer length processing algorithm built into a LabVIEW data acquisition and control interface, the multi-camera system is shown to provide assessments of transfer length within a nominal tolerance of +/− 1.5 inches using as few as six uniformly spaced surface strain measurements. This brings within reach the ultimate goal of providing the railroad tie manufacturer with the ability to measure the transfer length of every tie produced prior to leaving the plant, thereby ensuring that they are within an acceptable tolerance, and providing the means to quickly identify the need to modify production (e.g., concrete mix) if transfer length specifications fall out of desired range.

Wheel detection and axle counting systems provide the basic information for track vacancy detection, for controlling level crossing systems and a range of switching tasks or trigger functions. This technology is established throughout the world for standard gauge railways and is increasingly replacing track circuits. The requirements and the framework conditions in rail-based public transportation systems do, however, differ significantly from those in the standard-gauge railway and long distance track sectors. Nevertheless, in this segment too there is a clear trend towards wheel detection and axle counting systems. New lines and projects are increasingly being equipped with modern axle counting systems due to the benefits in terms of functionality and operating costs. Nowadays even in railed public transport systems, there is a clear move from track circuit systems to wheel detection and axle counting technology.

Commuter rail systems are widely regarded as an effective transportation alternative to reduce energy consumption and emissions in large urban areas. Use of commuter rail systems in the United States is on the rise, with annual ridership increasing by 28 percent between 1997 and 2007 [1]. With growing concerns about the sustainability and environmental impacts of transportation, modal energy efficiency is increasingly considered amongst the metrics to evaluate the benefits and costs of transportation systems and justify future investment. To gauge the relative long-term efficiency trends for rail as an urban transportation mode, this study analyzes historic trends in energy efficiency metrics for US commuter rail systems. Commuter rail systems receiving, or benefiting from, Federal Transit Administration (FTA) grants are required to report operations and energy consumption data on an annual basis to the National Transit Database (NTD). NTD data on energy consumption, operations, and services supplied from 1997 to 2011 are analyzed to determine historic trends in various energy efficiency metrics for the commuter rail mode as a whole. The data analysis and comparison of the results with the highway mode is complicated by the use of electric traction by some commuter rail operators. These operators report energy consumption in purchased electricity (kWh) instead of gallons of liquid fuel. The different approaches that can be employed to compare these two forms of propulsion and their intrinsic efficiencies and energy sources are discussed. Energy efficiency of each commuter rail system and its relationship to individual system characteristics during the study period are also analyzed. Finally, case studies of historic energy efficiency of individual commuter rail systems with longer operating histories and reporting data over the majority of the study period are contrasted against more recent start-up systems. While many systems outperform the energy efficiency of a typical light-duty vehicle, there are others that, due to a variety of system parameters and characteristics, fail to achieve a load factor great enough to make them more attractive than the highway mode on a gross average level. It is hoped that highlighting trends and variation in commuter rail energy efficiency will allow policymakers to make more informed decisions regarding the environmental benefits of rail as an urban transportation mode.

Automated in-plant diagnostic testing of prestressed concrete railroad crossties is now within reach due to recent progress in robust surface strain measurement techniques. The newly developed non-contact Laser Speckle Imaging (LSI) technique has been shown to provide rapid and accurate surface strain profile measurement, which is a key requirement for rapid transfer length assessment. Accurate determination of transfer length is critical for maintaining continuous production quality in the modern manufacture of prestressed concrete railroad crossties. Conventional assessment of transfer length generally presumes the underlying existence of a bilinear prestressing force distribution and a corresponding bilinear surface strain profile. Furthermore, it is well-known that this bilinear profile is smoothed due to the effects of finite gauge length during the process of measuring surface strain. In addition, recent extensive crosstie measurements in concrete railroad tie plants have shown significant departures from this simple bilinear profile, which bring to question the general validity and reliability of the traditional 95% AMS (95% Average Maximum Strain) method. Deviations from the simple bilinear profile shape were shown to be partially due to the non-prismatic shape of typical concrete railroad ties. In addition, extensive comparisons between predicted and measured surface strain profiles on numerous crossties suggest that the underlying strain distribution for crossties is best represented by an exponential strain profile, with an asymptotic approach to the fully-developed compressive strain. This is in contrast with extensive testing of prisms with fixed cross-section and fixed prestressing wire eccentricity, for which the surface strain appears to be best represented by the simple bilinear strain profile. Clearly, departures from non-prismatic behavior have added complexity to transfer length measurement. If accurate and reliable measurements of this important quality control parameter are to be realized, these issues of transfer length uncertainty need to be addressed. This paper provides an experimental comparison of several possible alternative transfer length assessment procedures, in an attempt to answer important uncertainty questions which need to be addressed if rapid real-time transfer length is to be achieved. It is shown that in spite of considerable differences in the transfer length processing methods, and significant departures from prismatic behavior, the averaged results are in many cases consistent with the simple bilinear underlying strain profile assumption. Bias in the measurement of crosstie transfer length due to non-prismatic behavior will also be investigated in this paper.

One of the most common failure modes of concrete crossties in North America is the degradation of the concrete surface at the crosstie rail seat, also known as rail seat deterioration (RSD). Loss of material beneath the rail can lead to wide gauge, rail cant deficiency, and an increased risk of rail rollover. Previous research conducted at the University of Illinois at Urbana-Champaign (UIUC) has identified five primary failure mechanisms: abrasion, crushing, freeze-thaw damage, hydro-abrasive erosion, and hydraulic pressure cracking. The magnitude and distribution of load applied to the rail seat affects four of these five mechanisms; therefore, it is important to understand the characteristics of the rail seat load distribution to effectively address RSD. As part of a larger study funded by the Federal Railroad Administration (FRA) aimed at improving concrete crossties and fastening systems, researchers at UIUC are attempting to characterize the loading environment at the rail seat using matrix-based tactile surface sensors (MBTSS). This instrumentation technology has been implemented in both laboratory and field experimentation, and has provided valuable insight into the distribution of a single load over consecutive crossties. A review of past research into RSD characteristics and failure mechanisms has been conducted to integrate data from field experimentation with existing knowledge, to further explore the role of the rail seat load distribution on RSD. The knowledge gained from this experimentation will be integrated with associated research conducted at UIUC to form the framework for a mechanistic design approach for concrete crossties and fastening systems.

The application of Doppler based, LI ght D etection A nd R anging (LIDAR or lidar) technology for measuring true ground speed in a non-contacting manner is investigated, as a replacement to wheel tachometers that are commonly used for such measurements. Measuring track speed and distance traveled is an essential part of rail geometry measurement systems. Wheel tachometer measurement accuracy can often be adversely affected by wheel vibrations, change in wheel diameter, and wheel slip in high traction conditions. LIDAR is a non-contact measurement device that uses the Doppler technology to accurately determine speed. Two LIDAR sensors are attached to the underside of a track geometry car with the sensors’ Class I laser beams facing the gauge corner of each rail. The LIDAR sensors measure the absolute ground speed for each rail, allowing for the determination of forward speed and track curvature. The results of the tests show high accuracy in LIDAR speed, distance, and track curvature measurements when compared with other conventional means that are used in the railroad industry and ground truth measurements. With further development, LIDAR sensors can replace wheel tachometers that are commonly used for speed and distance measurement, therefore eliminating the problems with mechanical reliability and the need for periodic calibration of wheel tachometers.

Knowledge of transfer length during production is critical for maintaining continuous production quality in the modern manufacture of prestressed concrete railroad crossties. Traditional laboratory methods for measuring transfer length, using manual instruments such as a Whittemore mechanical gauge or surface mounted resistance-type strain gauge, are simply not suitable for production operation. They are too time-consuming to implement, require extensive surface preparation, and can also require special operator training to provide accurate and reliable surface strain profiles from which the transfer length can be determined in a post-processing manner. In contrast with earlier manual methods, the newly developed non-contact Laser Speckle Imaging (LSI) technique has been shown to be capable of providing rapid and accurate surface strain measurement and consequently also rapid transfer length assessment. This system has recently been automated and combined with the new Zhao-Lee (ZL) least-squares strain profile fitting technique for quickly and reliably processing surface strain data. The automated system and processing procedure have been shown to provide an improved assessment of transfer length, unhampered by human intervention and subsequent potential human judgment bias. This paper presents recent progress toward the development of a 5-camera non-contact transfer length measurement system that is capable of continuous monitoring of railroad crossties in a production plant. This is made possible using an optimized version of the previously successful LSI system, which minimizes the number of surface strain measurements required to achieve reliable transfer length assessment. Experimental results and analysis will be presented for the latest multi-camera prototype concept for this new system design, demonstrating that only a few discrete surface strain measurements are required to achieve accurate and reliable transfer length assessment. Thus, for the first time it is now possible to envision practical real-time quality control monitoring of railroad crossties during an in-plant production operation.

This paper presents the detailed analysis of surface strain data obtained at six prestressed concrete tie plants in the United States. These data were obtained by the authors by conducting a total of 220 transfer length measurements on prestressed concrete railroad ties with different concrete-mix designs and reinforcement variations. The surface strain profiles of the railroad ties were obtained using the traditional Whittemore gage, as well as a rapid non-contact technology, called laser-speckle imaging (LSI), that was previously developed by the authors. The LSI technique achieved a microstrain resolution comparable to that was obtained using mechanical gauge technology. The measured surface strain profiles were then analyzed by both the 95% AMS (95% Average Maximum Strain) method, and the ZL (Zhao-Lee) method that was recently proposed by the authors. The ZL method is an unbiased statistical method that provides a more accurate and reliable transfer length determination. A direct comparison between the 95% AMS method and the ZL method was achieved by applying both methods to determine the 220 railroad tie transfer length values. The comparison confirmed the bias of the 95% AMS method in estimating transfer length value, as predicated by theoretical analysis.

One of the fundamental keys to improving track safety standards is to establish a strong correlation between track geometry variations and wheel-rail force parameters that are indicators of vehicle-track safety performance. In this study, wheel-rail forces were collected during field tests of a loaded lumber car and an empty tank car. Computer models of the two tested freight cars were built, and the models were calibrated using field test results. The computer models were then used to evaluate the impact of varying track geometry parameters on track safety using the maximum single wheel L/V ratio, maximum truck side L/V ratio, and minimum vertical wheel load ratio. It was confirmed again that the correlations between these force parameters and any individual geometry parameter were weak. With further investigation, it was found that much better correlation can be achieved using multiple regression techniques to define each wheel-rail force parameter as a function of all track geometry parameters combined together. Expressions of the maximum truck side L/V ratio, maximum single wheel L/V ratio, and minimum vertical wheel load ratio were obtained as functions of curvature, cross level, alignment, gauge, and cant deficiency using multiple regression analysis.

The aim of the paper is to classify restraining rail, discuss the advantages and disadvantages of each type of restraining rail, derive the formula to determine the flangeway gap, and finally, to suggest a type of restraining rail for use. Three types of restraining rail are classified as: 1. Active restraining rail: Defined as the restraining rail that reduces the angle of attack (AOA) by more than 50%. 2. Semi-active restraining rail: Defined as the restraining rail that reduces the AOA by 50% or less, preferably between 40% ∼ 50%. 3. Passive restraining rail: Defined as the restraining rail that does not reduce the AOA. In other words, it plays a passive role in steering the wheel. A design procedure is established to estimate the flangeway gap. The advantages and disadvantages of each type of restraining rail are discussed from design, maintenance, and functional points of view. The issue of optimization of the rail/wheel profile is also discussed in context with the presence of restraining rail. One component of the flangeway gap is the space required for the angularity of the wheel. 2D CAD drawings are not efficient for this purpose, as the drawings cannot consider the AOA and the height of restraining rail on top of the rail level. In this context, Nytram plot is a solution; however, the plot needs a number of sectional drawings at gauge point level, rail level, and the top-of-restraining-rail level from the 3D drawing. A mathematical model that counts both the AOA and the height of restraining rail from the top of the rail level is developed here to capture the essence of the Nytram plot, and thereby to assess the space required for the angularity of the wheel. Finally, a semi-active restraining rail, with a formula for flangeway gap, is suggested for use. Being less elaborate and less time consuming, the formula is easier and quicker than the Nytram plot to estimate flangeway gap. Moreover, one can quickly assess the effect of wheel size, the height of the restraining rail from the top of rail, and the radius of the curve on the flangeway gap.