Resumen de Study and characterisation of surface integrity modification after ultrasonic vibration-assisted ball burnishing
This dissertation is an experimental research project into the mechanical effects of the ultrasonic vibration-assisted ball burnishing process on the surface integrity of surfaces machined through ball-end milling. Due to the lack of commercial tools able to perform this process, the study includes firstly the design and characterisation of a prototype to that effect. An experimental analysis is then undertaken, applying the process to AISI 1038 and Ti-6Al-4V surfaces of high industrial and aeronautical value. The experimental campaign is designed based on a Taguchi orthogonal array that includes five factors, namely: preload, number of passes, feed velocity, strategy and initial surface texture. Results are analysed in terms of topological characteristics, residual stress and hardness, in order to identify and understand the impact of process parameters on surface integrity, to define the best parameters for performing the process and to assess the positive effects caused by the introduction of vibrations as a means of assistance. Results reveal that the initial texture is the most influential parameter on all outcomes. Texture results show that the vibrations can enhance the roughness and texture results, as long as they have sufficient low initial amplitude. Furthermore, only the preload and number of passes influence the results, with a pair of values being found in all cases that serve a threshold from which further plastic strain is detrimental for the final surface topology. In terms of residual stress, all parameters are influential in the results, especially the burnishing strategy, through which a certain component of the residual stress tensor can be adequately reinforced. Finally, the burnishing operation proves to modify the hardness of deep layers down to 0.5 mm, applying the vibration-assisted process. The main conclusion is that the optimal parameters for performing the process are different with regards to the optimisation objective. Some useful combinations are proposed for performing the process depending on the desired target.
