Rail fatigue damage and wear have a major financial and safety impact on rail-roads. Since 2017, the authors have created and shared a number of highly-detailed rail loading environments aimed at testing and calibrating rolling contact fatigue damage models and wear models. Creating these loading environments involves the collection of real-world data – track geometry, car types, car dimensions, truck suspension characteristics, car weights, train speeds, friction conditions, wheel and rail transversal profiles, rail surface hardness, rail crack depth, tonnage since last grinding, and rail surface photographs – repeated at each site several times over several months. Contact conditions are then computed using a large number of dynamic simulations with stochastically-selected inputs to best represent the actual traffic going through each site. Simulation outputs include contact patch characteristics – centre of contact, forces, creepages, traction forces and contact patch geometry – as well as the position of the car body, trucks, and axles relative to the track. The authors use pummeling analyses to validate the results of these simulations by comparison with existing rail surface conditions. With each iteration, the number and variety of test sites increases and the methodology evolves and provides more detailed datasets.