Much research has been performed on predicting the derailment of a rail car when properly situated on rails. However, less work has been performed on evaluating what happens to a railcar’s wheel after derailment. Recent work has used a simplified off-track ground friction coefficient to aid in the prediction of post-derailment car trajectories. This paper uses soil cutting theory to investigate the lateral and longitudinal forces acting on a derailed wheel in different soil types. Findings are compared to the use of the simplified off-track ground friction and may improve the accuracy of current derailment simulation techniques.

Predicted wheel performance is an invaluable tool in developing new components such as wheel profiles and truck components and also in understanding and controlling wheel and rail wear and rolling contact fatigue (RCF). This paper outlines a vehicle dynamics trial using VAMPIRE Pro to compare the predicted wear and RCF performance of the WRISA2 wheel profile developed by the National Research Council of Canada (NRC) and the United Kingdom P8 wheel profile using measured wheel profiles from an in-service trial. WRISA2 and P8 profiles were fitted to two passenger trains running in normal service. Wheel profiles were measured every 10,000 miles. These measured profiles were used to predict wear and RCF damage for each wheel of the investigated rail vehicle, using a combination of VAMPIRE transient analysis and another program called the “Whole Life Rail Model” (WLRM). This process was repeated up to 190,000 miles run in service, allowing a clear comparison of the changing rail wear and RCF performance of the two profiles up to this mileage. This process was automated using new features within VAMPIRE that allow communication to 3 rd party computer programs including the WLRM, Microsoft Excel, and Microsoft Visual Basic. This research process presents itself to be a very useful tool in predicting wheel wear performance for any number of new wheel and truck components.

A predictive life tool for case-carburized railcar tapered roller bearings, called the Service Load Factor (SLF), makes use of residual stress and retained austenite data to assess the condition of bearings as a result of service history. It is well known that the interaction between retained austenite transformation and the development of residual stresses are primary factors contributing to cone bore growth and decreased resistance to raceway spalling. This paper attempts to quantify these changes throughout the life of a bearing to provide insight into the predicted remaining service life of the bearing. Also, this paper will cover the use of the SLF as a failure analysis tool in order to quantify bearing damage due to shifted lading or worn adapters.

Screw spikes, also known as coach screws, are an advanced alternative to common cut spikes for track fastening. Despite their ability to secure tie plates with a clamp load and utilization of high strength steels, they are still susceptible to bending fatigue failure from lateral wheel loads. A novel method of measuring these bending loads on screw spikes was developed and implemented to characterize the load environment of the screw spikes. Results indicated that measured peak bending loads under lateral wheel loads reached as high as 10,000 lbs for individual spikes, while others carried no load whatsoever. A finite element model was developed to determine the tensile stress fields created by the measured bending loads. A good correlation was found between the FEA model predicted point of highest stress and the location of fracture. Through the testing and analysis it was determined that lateral wheel loads are not distributed evenly among the four screw spikes of a single tie plate. Instead, it was found that one spike carried nearly no load while the spike opposite of it carried more load. Using the finite element analysis it was determined that the spike exposed to the higher loading was subjected to tensile stresses above its endurance limit, which would eventually lead to a bending fatigue failure.

Mechanical damage to tapered roller bearings resulting from wheel impacts is a growing concern in the railroad industry. Repeated impacts can cause plastic deformation, wear, and eventually fatigue in roller bearing cages. A servo-hydraulic system was implemented in order to impact a bearing assembly with 65 g peak acceleration impacts at 6 Hz. Three different cage designs were tested for the Class F and K bearings: the standard steel design, and two polyamide cage designs. Testing was conducted on the three cage designs in impact increments of 10 2 , 10 3 , 10 4 , 10 5 , 10 6 , and 15×10 6 cycles. Between each testing increment, cages were removed from a special two-piece cone to inspect damage. Mechanical damage in the forms of plastic deformation and wear were measured by means of their area and depth. Results indicated that although high acceleration impacts were inflicted on the bearing; none of the cages developed fatigue fractures in the roller pocket corners consistent with field failures. However, a large amount of hourglass shaped damage in the roller pockets was recorded. Comparing the three designs revealed that although the polyamide cage designs were lower strength, they were more resistant to plastic deformation and wear than the steel cage design.

Wheel flats are common railcar wheel defects that cause bearing failures and train derailments. In an effort to better understand the impacts created by a wheel flat, a nonlinear finite element model using LS-DYNA was created of a single 914.4 mm (36 in) railcar wheel with a wheel flat and a 3.24 m (127.5 in) half track with one rail, six wood ties and tie plates. Discrete element springs and dampers simulated the ballast and subgrade. Goals of this model were to accurately simulate a wheel flat impact at varying speeds and to monitor the plastic deformation that occurs at the sharp edges of the wheel flat during rotation. Producing and validating this model then could be used to test modifications to the track and wheel to lower the severity of wheel impacts. Results indicated an accurate simulation; however, improvements with the material properties, suspension model and mesh sensitivity can be made.