The Gautrain Rapid Rail Link (GRRL) is a rail transit system in South Africa that links Johannesburg and Pretoria, as well as Johannesburg and the O.R. Tambo International Airport. Travelling at speeds of up to 160 km/h, the Gautrain system is the first of its kind on the African continent. This dissertation covers an investigation into the relationship between the vertical and lateral forces, speed and superelevation in a GRRL curve. The research described in this dissertation is based on an experiment which involved running a test train through a curve at various speeds, changing the cant of the curve by tamping and repeating the train runs. The cant was changed due to high wheel wear rates. The curve already had a cant deficiency, and this cant deficiency was subsequently increased by reducing the curve’s cant. Assessing the before and after tamping test data validated the existence of the expected relationships between the vertical and lateral rail forces, the speed and the cant. The change in cant had a minimal effect on the magnitude of the vertical forces, although a transfer of loading between the high and low legs did occur. The theory indicates that the 14 % reduction in cant in this curve given all of the other curve characteristics should have resulted in an increase in the lateral forces. There was however a roughly 50 % reduction in the maximum lateral forces after the cant was reduced that can be explained from a train dynamics point of view. In addition, there was an increase in safety due to a reduced derailment ratio at this curve’s normal operating speed of 85 km/h. It is not unreasonable to presume that a 50 % reduction in the maximum lateral forces could lead to a halving of the wear rate of the rail and wheels in this curve with similar results to be expected in other curves on the rail network. / Dissertation (MEng)--University of Pretoria, 2016. / The Chair in Railway Engineering at the University of Pretoria / Civil Engineering / MEng / Unrestricted

Heavy haul railway operations permit the transport of huge volumes at lower cost than other modes of transport. The low cost can only be sustained if the maintenance costs associated with such railway operations are minimised. The maintenance costs are mainly driven by wheel and rail damage in the form of wear and rolling contact fatigue (RCF). Low wear rates in the wheel-rail interface have resulted in an increase in the prevalence of rail RCF, thereby increasing rail maintenance costs. The aim of this study is to develop an approach to reduce rail RCF on South Africa’s iron ore export line by managing the worn wheel shape. This approach is developed by evaluating wheel and rail profile shapes that contribute the most to RCF initiation, studying the influence of suspension stiffness and rail profile changes as well as a redesign of the wheel profile. The influence of wheel and rail profile shape features on the initiation of rolling contact fatigue (RCF) cracks was evaluated based on the results of multibody vehicle dynamics simulations. The damage index and surface fatigue index were used as two damage parameters to assess the influence of the different features. The damage parameters showed good agreement to one another and to in-field observations. The wheel and rail profile shape features showed a correlation to the predicted RCF damage. The RCF damage proved to be most sensitive to the position of hollow wear and thus bogie tracking. RCF initiation and crack growth can be reduced by eliminating unwanted shape features through maintenance and design and by improving bogie tracking. Two potential mitigation measures had been adapted from those published in literature to reduce RCF. The mitigation measures involved changes in suspension stiffness to spread wheel wear across the tread and the use of gauge corner relief rail profiles. These mitigation measures were evaluated by means of multibody dynamics and wear maintenance costs. These mitigation measures, however, did not prove to be successful in reducing RCF initiation while maintaining a low wheel wear rate. The current operating conditions on South Africa’s iron ore line, although still not optimal overall, were found to be better in terms of their wear and RCF performance than the two proposed RCF mitigation measures. Based on the finding of the study on two RCF mitigation measures it was recommended that a conformal wheel profile be developed to spread the wheel wear across the tread to reduce the occurrence and propagation of RCF cracks, while still maintaining low wheel wear rates. A comparative study of this new wheel profile design and the current wheel profile design was therefore performed using multibody dynamics simulation together with numerical wheel wear and RCF predictions. The advantages of the conformal wheel profile design were illustrated by evaluating the worn shape and resulting kinematic behaviour of the conformal design. The conformal design had a steadier equivalent conicity progression and a smaller conicity range compared with the current wheel profile design over the wheel’s wear life. The combination of a conformal wheel profile design with 2 mm hollow wear and inadequate adherence to grinding tolerances often result in two-point contact, thereby increasing the probability of RCF initiation. The conformal wheel profile design was shown to have many wear and RCF benefits compared with the current wheel profile design. However, implementation of such a conformal wheel profile must be accompanied by improved rail grinding practices to ensure rail profile compliance. Based on these findings an approach is proposed where the conformal wheel profile design together with improved compliance of the in-service rail profiles to the target rail profile are implemented. This has the potential to reduce RCF initiation on South Africa’s iron ore export line. / Thesis (PhD)--University of Pretoria, 2017. / Mechanical and Aeronautical Engineering / PhD / Unrestricted

This study provides a detailed experimental evaluation of wheel-rail interaction for railroad vehicles, using the Virginia Tech Federal Railroad Administration (VT-FRA) Roller Rig. Various contact dynamics that emulate field application of railroad wheels on tracks are set up on the rig under precise, highly-controlled and repeatable conditions. For each setup, the longitudinal and lateral traction (creep) forces are measured for different percent creepages, wheel loads, and angles of attack. The tests are performed using quarter-scaled wheels with different profiles, one cylindrical and the other AAR-1B with a 1:20 taper. Beyond the contact forces, the wheel wear and the deposition of worn materials are measured and estimated as a function of time using a micron-precision laser optics measurement device. The change in traction versus amount of worn material at the contact surface is analyzed and related to wheel-rail friction. It is determined that the accumulation of the worn material at the contact surface, which appears as a fine gray powder, acts as a friction modifier that increases friction. The friction (traction) increase occurs asymptotically. Initially, it increases rapidly with time (and worn material accumulation) and eventually reaches a plateau that defines the maximum friction (traction) at a stable rate. It is estimated that the maximum is reached when the running surface is saturated with the worn material. Prior to the saturation, the friction increases directly with an increasing amount of deposited material. The material that accumulates naturally at the surface—hence, referred to as "natural third-body layer"—is estimated to be a ferrous oxide. It has an opposite effect from the Top of Rail (ToR) friction modifiers that are deposited onto the rail surface to reduce friction in a controlled manner. Additionally, the results of the study indicate that longitudinal traction decreases nonlinearly with increasing angle of attack (AoA), while lateral traction increases, also nonlinearly. The AoA is varied from -2.0 to 2.0 degrees, representing a right- and left-hand curve. Lateral traction increases at a high rate with increasing AoA between 0.0 – 0.5 degrees, and increases at a slow rate beyond 0.5 degree. Similarly, longitudinal traction reduces at a high rate for smaller AoA and at a slower rate for larger AoA. For the tapered wheel, an offset in lateral forces is observed for a right-hand curve versus a left-hand curve. The wheel taper generates a lateral traction that is present at all times. In one direction, it adds to the lateral traction due to the AoA, while in the opposite direction, it subtracts from it, resulting in unequal lateral traction for the same AoA in a right-hand versus a left-hand curve. The change in traction with changing wheel load is nearly linear under steady state conditions. Increasing the wheel load increases both longitudinal and lateral tractions linearly. This is attributed to the friction-like behavior of longitudinal and lateral tractions. An attempt is made to measure the contact shape with wheel load using pressure-sensitive films with various degrees of sensitivity. Additionally, the mathematical modeling of the wheel-roller contact in both pure steel-to-steel contact and in the presence of pressure-sensitive films is presented. The modeling results are in good agreement with the measurements, indicating that the pressure-sensitive films have a measurable effect on the shape and contact patch pressure distribution, as compared with steel-to-steel. / Doctor of Philosophy / This study provides a detailed experimental evaluation of wheel-rail interaction for railroad vehicles, using the Virginia Tech Federal Railroad Administration (VT-FRA) Roller Rig. Better understanding the dynamics and mechanics of wheel-rail interaction would significantly contribute to the development of technologies, materials, and operational methods that can further improve fuel efficiency, and reduce wheel and rail wear. Considering that the railroads are the backbone of cargo and passenger transportation and are critical to economic well-being, the results of this study are expected to contribute to the betterment of society. An attempt is made to emulate the field application of railroad wheels on tracks on the rig under precise, highly-controlled and repeatable conditions. For each set up, the contact forces are measured for different parameters, such as wheel loads. Beyond the contact forces, the wheel profile degradation and the deposition of worn materials are measured and estimated as a function of time using a micron-precision laser optics measurement device. It is determined that the accumulation of the worn material at the contact surface, which appears as a fine gray powder, increases contact forces. The effect of wheel load on contact forces is almost linear. Additionally, the results of the study indicate that the yaw angle between the wheel and the roller (AoA) changes the contact forces direction, which has a higher rate of change for a small AoA such as 0.0 – 0.5 degrees, compared to a larger AoA. An attempt is made to measure the contact shape with wheel load and AoA using pressure-sensitive films with various degrees of sensitivity. Additionally, the mathematical modeling of the wheel-roller contact in both pure steel-steel contact and in the presence of pressure-sensitive films is presented. As expected, both the model and test result indicate that the presence of a film at the contact surface changes both the dimensions and pressure distribution at the contact patch. Quantifying the distortion that occurs as a result of the pressure-sensitive film allows for a better assessment of the pressure distribution measurements that are made with the films in order to potentially discount the resulting distortions.

Road vehicle tyres, railway wheels and ball bearings all generate rolling contact forces which are transferred within a finite area of contact between the rolling element and the substrate. Either it is visible or not for the human eye, a certain degree of roughness is always present on the contacting surfaces and it influences the generation of both vertical and lateral contactforces. The purpose of this investigation is to enhance the understanding and modelling of the influence from small-scale surface roughness on the generation of rolling contact forces. To this end, a computationally efficient method to include roughness-induced contact nonlinearities in the dynamic modelling of rolling contacts is proposed. The method is implemented in a time domain model for vertical wheel–track interaction to model rolling-induced rail vibrations, showing good agreement with measurements. Furthermore, a test rig is developed and used for the investigation of tyre–road rolling contact forces. Detailed studies are performed on the influence of substrate roughness on the resulting contact forces for a tyre tread block which is rolling at different operating conditions. The choice of substrate as well as the rolling velocity and the slip ratio is observed to have significant influence on the resulting friction coefficient. For high slip ratios, stick–slip oscillations appear, exhibiting frequency content which is largely dependent on the choice of substrate. The outcomes of this study can potentially be used to improve future tyre–road contacts with respect to wear, traction and noise generation. / <p>QC 20161013</p> / Centre for Eco2 Vehicle Design