The objective of this work was to provide a new methodology to improve thermite welds. Improving thermite weld is important because it is a major concern for the North American railways and it is crucial to meet the network demands in terms of safety, reliability and ridership operations. More specifically, the objective of this work was to improve fatigue performance (i.e., strength when subjected to cyclic stress) by minimizing the two most common types of porosity -- gas and shrinkage. Gas porosity is the result of the precipitation of oxygen during solidification, and shrinkage pores are cause by thermal contraction combined with lack of melt feeding into the solidifying casting, which in this case is the thermite weld. To accomplish this objective, we have proposed a vibration technology applied during the solidification of the thermite welds to reduce the gas pores and refine the microstructure. The result is an improved weld material with superior mechanical properties. The work was divided into three major stages: i) design of experiments and thermal weld casting, ii) mechanical testing, optical microscopy, and iii) analysis and reporting of results. Due to the expense associated with performing first-time experiments on numerous thermite welds, we used a statistical method, Design of Experiments (DOE), to efficiently explore the process variables of our vibration system. The purpose of DOE is to determine the specific variable that control a multi-variable process. In addition, a rail weld simulation device was developed to use in the laboratory to enable pre-test the vibration device and explore the controlling variables. This information allows for a more efficient exploration of the vibration process when applied to thermite welds. The operational variables considered for the design of experiments were vibration frequency, vibration force, vibration orientation and the distance of the vibrator from the weld. The first set of experiments was conducted in our laboratory scale simulator to understand the fluid mechanics of the vibration on a liquid. The experiments were performed using water containing marker. The simulation results permitted the identification of the most important operating parameters to be tested in 16 experimental welds. Three additional welds were cast following current standard practices and used as controls. The results indicate that the treatment is effective and improves the service life of the welds. Results from thermite welds treated with our vibration process show improvements in strength of up to 12% and a reduction in porosity. The porosity reduction has an expected improvement on the fatigue endurance limit of up to 30%. The treated welds were inspected using the protocols and recommended practices in American Railway Engineering and Maintenance-of-Way Association (AREMA). These tests demonstrated that the vibration treatment is suitable for revenue service use as it does not have negative effects on welds. The vibration parameter that has the greatest benefit is frequency, followed by the vibrating force. The other two parameters, orientation and distance between the vibrator and the weld, have lesser influence. This report summarizes the methodology followed in order to propose a successful weld treatment.

The final report is available.